The gadolinium ion, a rare earth lanthanide series metallic salt, is the active, favorable center for chelation to polyaminocarboxycilic acid ligands in GBCAs.1–5 Seven unpaired electrons with isotropic distribution (in the 4f shell) offer an electrically charged magnetic center that has the strongest effect of all elements on electronic T1 relaxivity for local water molecules.1–3 These 7 unpaired electrons and 9 coordination sites result in the large paramagnetic susceptibility of gadolinium.1,6 The GBCAs are classified on the basis of charge and structure (which depends on the configuration of the chelated ligand around the gadolinium ion). Structure and charge both contribute to the stability of these agents.7,8 Clinically, GBCAs are administered IV for contrast-enhanced MRI in humans and domestic animals, which results in improved diagnostic capability and lesion conspicuity, especially for lesions that might otherwise be difficult to detect.2 Typical doses for humans and domestic animals range from 0.1 to 0.2 mmol/kg for magnet strengths > 0.5 T.7 Historically, GBCAs have been considered safe for use, despite the cytotoxic effects of the gadolinium ion.7
Persistent, symmetric T1 signal hyperintensity of the dentate nucleus is seen in mice and humans following multiple (≥ 6) IV doses of GBCAs and is suggestive of deposition of gadolinium within brain tissue.7,9,10 In mice, T1 signal hyperintensity of the deep cerebellar nucleus and gadolinium deposits in the cerebellum have been detected after multiple high doses (20 doses at 0.6 mmol/kg, IV) of gadodiamide.9,11 In contrast, similar doses of gadoterate meglumine, an ionic macrocyclic agent, do not result in persistent changes in signal intensity.9,11 Cerebellar, cerebral, and subcortical retention of gadolinium measured with ICP-MS was significantly higher in rats given gadodiamide, compared with rats given gadoterate meglumine.9 When administered into the ventricular system, both agents cause focal and generalized myoclonus for several hours, with the cerebellar region having the greatest sensitivity to gadodiamide.12 When administered IV, gadopentetate dimeglumine at doses of 0.1 and 0.2 mmol/kg, respectively, significantly increases seizure frequency in 50% and 75% of dogs with osmotically induced disruption of the blood-brain barrier.13
Persistent regional brain enhancement occurs in both juvenile and adult human patients with normal renal function, primarily after multiple (> 2 to 12) doses (0.1 mmol/kg) of gadodiamide and gadopentetate dimeglumine.7,11,14–20 Gadolinium deposition in the brain has been detected in deceased human patients with normal renal and hepatic function.3,16,21 In humans, a relationship may exist between neural gadolinium deposition and increased clinical disability, lesion load, brain atrophy, or brain irradiation.3,11,22–25 Cadaveric brain samples of 13 people that received gadodiamide IV as a part of ≥ 4 contrast-enhanced MRI evaluations between 2000 and 2014 had dose-dependent retention of gadolinium in the brain independent of renal function.3,7 Gadolinium deposition in the brain has been detected after multiple exposures to various combinations of gadopentetate dimeglumine and gadoteridol.3,7 Gadolinium deposition (both linear and macrocyclic agents) in the globus pallidus, dentate nucleus, and bone has been documented in 9 patients that underwent contrast-enhanced MRI because of various illnesses23; median bone deposition was 23 times as great as that of the brain. Single or multiple administrations of both macrocyclic nonionic agents and linear nonionic agents result in deposition in the brain.23 Currently, it is not possible to determine distribution characteristics and the underlying reasons for patterns of deposition.8,23
The perception that chelation completely eliminates toxic effects of gadolinium ions has been questioned. The retention of GBCAs within the brain may pose a dilemma for physicians who rely on contrast-enhanced MRI to obtain better MRI specificity. The National Center for Toxicological Research of the US FDA issued a GBCA safety announcement26 that prompted investigations to determine mechanisms of gadolinium retention and risks posed with regard to gadolinium deposition in the brain after single or multiple administrations in people.1,3,21,24 The European Medicines Agency also reviewed the risk of gadolinium deposition in the brains of patients undergoing contrast-enhanced MRI evaluations.27 The FDA and National Institutes of Health have advised that use of GBCAs be limited to necessary clinical circumstances, that the need for repetitive use in MRI evaluations for well-established protocols be reassessed, and that the National Institutes of Health recommendations for clinical use be heeded until further evaluation of risks can be investigated.24,26 Subsequently, the FDA issued a statement28 that evidence does not exist to support the contention that gadolinium retention is harmful; however, the FDA recommendations to health-care providers regarding use of GBCAs remained unchanged. Currently, according to the FDA, no clinical signs or syndromes can be reasonably linked to gadolinium deposition.26,28
Dogs are a historically viable means for translational research of brain physiology and disease states.29,30 The clinical relevance of residual gadolinium in brain tissue of dogs is unknown. Therefore, goals of the study reported here were to determine whether intracranial gadolinium retention occurs in young healthy dogs administered various doses (including several subclinical doses) of gadodiamide (a linear nonionic contrast agent) as a single IV injection, to determine patterns of gadolinium retention and affinity, and to provide information regarding the short-term effect of fixation (formalin-fixed tissue compared with fresh-frozen tissue) on gadolinium tissue concentration in the brains of dogs that received gadodiamide.
Materials and Methods
Animals
Fourteen healthy purpose-bred hound dogs were included in the study. Physical (performed by an institutional laboratory animal veterinarian and a diagnostic imaging resident [AML]) and neurologic (performed by a board-certified veterinary neurologist) examinations as well as a CBC,a serum biochemical analysis,b and serum occult heartworm antigen testc were performed on all dogs. Design and execution of in vivo portions of the study were approved by an institutional ethical review board (Institutional Animal Care and Use Committee at Mississippi State University [Nos. 15012, 12065, and 15075]).
Dogs were allocated into 3 groups. Additional testing for dogs of group 1 (n = 8) included determination of serum titers for infectious diseasesd (Ehrlichia canis, Rickettsia rickettsii, Borrelia burgdorferi, and Babesia canis), MRI of the brain, and analysis of a sample of CSF. Cytologic examination of CSF and titer determination of CSF to detect West Nile virus were performed at a veterinary diagnostic laboratory.e For dogs of groups 2 (n = 5) and 3 (1), MRI of the brain, testing to determine serum titers for infectious diseases, and examination of CSF were not performed. All testing was performed 1 week (group 1) or 8 weeks (groups 2 and 3; median, 1 day; mean, 24.6 days; range, 1 to 54 days) before exposure to the GBCA.
Dogs were excluded from the study if they had abnormal results for the neurologic examination, CBC, blood biochemical analysis, or urinalysis; were seropositive for infectious diseases; had positive results for the occult heartworm antigen test; or had abnormal findings for MRI of the brain. Dogs also were excluded if they had complications related to anesthetic and diagnostic imaging procedures, collection of CSF, surgical interventions, or euthanasia; had an adverse event after IV administration of the GBCA; or had a history of IV administration of a GBCA.
GBCA administration
Group 1 dogs received a single IV injection of gadodiamidef at one of the following doses: 0.006 mmol/kg (1 dog), 0.0125 mmol/kg (1 dog), 0.025 mmol/kg (2 dogs), 0.05 mmol/kg (2 dogs), or 0.1 mmol/kg (2 dogs). Group 2 dogs received a single IV injection of gadodiamide at one of the following doses: 0.025 mmol/kg (1 dog), 0.05 mmol/kg (2 dogs), or 0.1 mmol/kg (2 dogs). The single dog of group 3 (control dog) did not receive an injection of gadodiamide. Instead, the dog received a single IV injection of a placebo (3 mL of saline [0.9% NaCl] solutiong). All IV injections were administered into the right cephalic vein, which was followed by IV administration of saline solutiong (5 mL/kg/h).
After gadodiamide was injected, dogs were used in a surgical teaching laboratory. Gastrointestinal and orthopedic surgeries, which did not involve disruption of the blood-brain barrier, were performed. At the completion of the surgeries, the anesthetized dogs were euthanatized by injection of euthanasia solutionh (4.5 mL/kg).
Tissue processing and histologic examination
Necropsy was performed on all 14 dogs. Brain harvesting and tissue sectioning were performed by a board-certified veterinary pathologist (AJC). Transverse samples for histologic examination were collected from each of the following regions of all dogs: parietal lobe, piriform cortex, thalamus, cerebellum, frontal lobe white matter, brainstem, motor cortex, and hippocampus. Tissues were processed routinely, embedded in paraffin, sectioned at a thickness of 4 μm, and stained with H&E stain for microscopic examination. Tissue samples (1 cm3) of the frontal lobe white matter, parietal lobe, piriform cortex, thalamus, cerebellum, and brainstem were collected from both hemispheres. Samples from one of the hemispheres were fixed by immersion in neutral-buffered 10% formalin for 22 to 278 days. Thin slices of fresh tissue samples from the contralateral hemisphere were frozen at −80°C. Additional samples were collected from the frontal lobe white matter, parietal lobe, piriform cortex, thalamus, cerebellum, and brainstem and placed in Karnovsky fixative (2% paraformaldehyde and 2.5% glutaraldehyde in 0.1M sodium cacodylate buffer; pH, 7.2). Formalin-fixed tissue samples for dogs of group 1 and fresh-frozen and formalin-fixed tissue samples for dogs of groups 2 and 3 were submitted for ICP-MS.
Gadolinium analysis
Tissue samples were analyzed and residual gadolinium was quantified with ICP-MS by a board-certified veterinary toxicologist (JPB). Selected regions were trimmed from fixed tissue; an industrial wipei was used to remove excess moisture from the sample, and the sample was placed in sterile transfer tubes. The aforementioned samples from all dogs were labeled, stored, and refrigerated in individual tissue vials during shipment. Formalin-fixed samples (1 cm3) from the frontal lobe white matter, parietal lobe, piriform cortex, thalamus, cerebellum, and brainstem of each dog were dried for approximately 20 hours at 95°C in a gravity convection oven.j Laboratory personnel were unaware of the gadodiamide dose for each sample. Each standard solution (200 μL), spiked control sample, and unknown sample were pipetted and diluted with 5 mL of a solution containing 0.5% EDTA and Triton X-100, 1% ammonia hydroxide, 2% propanol, and internal standards (scandium, rhodium, indium, and bismuth; 20 μg/kg). The mass spectrometerk was adjusted to yield a minimum sensitivity of 6,000 counts/s for yttrium (1 μg/kg; mass, 89), < 1.0% oxide as determined by the 156-to-140 mass ratio, and < 2.0% double-charged ions as determined by the 70-to-140 mass ratio. Gadolinium was calibrated by use of a 5-point linear relationship of the analyte-to-internal standard response ratio. Bismuth-209 was used as an internal standard. Helium was used as a collision gas to control polyatomic interferences with gadolinium. Gadolinium and the internal standard mixture were obtained from a manufacturing facility.l The mass spectrometer was calibrated for gadolinium concentrations between 0.3 and 200 ng/g. Gadolinium tissue concentrations were determined by multiplication of the weight of gadolinium per milliliter in the digested solution by the dilution factor and then division of that value by the tissue sample weight. Limit of detection for the gadolinium analysis was 0.00012 μg/kg.
Electron microscopy with electron probe microanalysis
Scanning electron microscopy was performed in conjunction with TEM and electron probe microanalysis at the Institute for Imaging and Analytical Technologies to characterize and quantify the distribution of gadolinium deposits. Samples were fixed in neutral-buffered 10% formalin as previously described. Small formalin-fixed samples (approx 1 mm2 for TEM and approx 1 cm for SEM) were excised and placed into half-strength Karnovsky fixative at 4°C until further processing. Samples then were rinsed and dehydrated in a graded series of ethanol solutions.
The TEM samples were infiltrated with Spurr resin. Ultrathin (approx thickness, 80 nm) sections were cut with an ultramicrotomem and collected on 75-mesh coated copper grids. Sections were stained with lead citrate and initially examined by use of a TEMn at 80 kV for image generation and orientation of potential gadolinium locations. Once key areas were identified, grids were reexamined at 200 kV with a TEMo equipped with an energy-dispersive spectroscopy systemp for elemental analysis.
Samples for SEM were chemically dried with hexamethyldisilazane and then air-dried. Dried samples were affixed to aluminum stubs with double-sided carbon tape and examined in an SEMq at a 10-kV and 40-pA current by use of the back-scattered detector for imaging and an energy-dispersive spectroscopy systemr for elemental analysis.
Statistical analysis
To determine the effect of dose on gadolinium retention in the brain, data were assessed by use of linear mixed-effects modelss by an investigator who was a board-certified veterinary epidemiologist (RWW). For formalin-fixed tissues, separate analyses were conducted for group 1, group 2, and groups 1 and 2 combined. For analyses of an individual group (ie, group 1 or group 2), dose, region, and the dose-by-region interaction were fixed effects in the model. Dose, group, region, and the dose-by-group, dose-by-region, and group-by-region interaction were independent variables for analyses of the combined groups (group 1 and 2 combined). Another analysis was conducted for group 2 to compare fresh-frozen and formalin-fixed tissues. Dose, region, tissue type (fresh-frozen or formalin-fixed tissues), and all 2-way interactions were fixed effects in the model.
For all models, dose was considered a continuous variable and was included as a covariate. Dog was included as a random effect in all models. When interaction terms were not significant, they were sequentially removed (the term with the largest P value was removed, and the model was refit). Differences in least squares means with an adjustment for multiple comparisonst were determined for significant main effects or interaction terms. Diagnostic plots of gadolinium residues for each outcome were assessed to ensure that assumptions of the statistical method had been met. A value of P < 0.05 was used to determine significant differences.
Results
Animals
Fourteen healthy purpose-bred dogs of similar age (median, 8.5 months; mean, 8.6 months; range, 5 to 13 months) and body weight (median, 23.75 kg; mean, 23.76 kg; range, 17.5 to 32.0 kg) met the inclusion criteria and were included in the study. No dogs were excluded, and none had a history of illness. Five dogs were sexually intact males, 8 were sexually intact females, and 1 was a spayed female. Dogs were presumed to have a complete and intact blood-brain barrier at the time of in vivo exposure to a single standard (0.1 mmol/kg) dose or fractional dose of gadodiamide prior to euthanasia. Dogs of group 1 received gadodiamide 3 to 7 days (median, 5.4 days; mean, 5 days) before euthanasia, whereas dogs of groups 2 and 3 received gadodiamide or saline solution, respectively, 8.3 to 8.9 hours (median, 8.7 hours; mean, 8.6 hours) before euthanasia. Dogs underwent all interventions without complications.
Tissue processing, tissue storage, and histologic examination
Brains were harvested 1 to 3.5 hours after dogs were euthanized and were kept on ice until processing. Formalin-fixed brain samples from dogs of group 1 and fresh-frozen and formalin-fixed brain samples from dogs of groups 2 and 3 were archived in our institutional biospecimen repository. Brain samples of group 1 dogs were stored in formalin for 150 to 278 days (median, 157.5 days; mean, 200 days), whereas brain samples of all group 2 dogs and the group 3 dog were stored in formalin for 22 days (prior to TEM) and 69 days (prior to ICP-MS). Median time in Karnovsky fixative prior to SEM and TEM was 36 hours. Formalin-fixed brain samples from dogs with the highest retention of gadolinium (dogs that received 0.5 or 0.1 mmol/kg) were submitted for SEM (group 1) and TEM (groups 1 and 2) for gadolinium identification. All dogs had grossly and histologically normal brain tissue; there was no evidence of associated or underlying pathological conditions detected with light microscopy.
Effect of gadolinium exposure on tissue retention
Gadolinium retained in the brains of the healthy dogs was quantified with ICP-MS (Table 1). There was elemental gadolinium retention in all dogs of groups 1 and 2 regardless of the dose administered, including the lowest fractional dose (0.006 mmol/kg; Figure 1). Intracranial gadolinium retention was detected in the 6 regions, with concentrations ranging from 1.7 to 162.5 ng of gadolinium/g of brain tissue for group 1 and from 67.3 to 1,216.4 ng of gadolinium/g of brain tissue for group 2. The highest mean gadolinium concentrations were detected in the cerebellum (mean, 191.16 ng of gadolinium/g of brain tissue; range, 10.92 to 658.9 ng of gadolinium/g of brain tissue), parietal lobe (mean, 221.7 ng of gadolinium/g of brain tissue; range, 7.88 to 910.30 ng of gadolinium/g of brain tissue), and piriform cortex (mean, 191.9 ng of gadolinium/g of brain tissue; range, 15.47 to 839.0 ng of gadolinium/g of brain tissue). Despite receiving no gadodiamide, the dog of group 3 had 5.3, 3.2, 0, 3.1, 2.4, and 1.4 ng of gadolinium/g of brain tissue in the brainstem, cerebellum, frontal lobe, parietal lobe, piriform cortex, and thalamus, respectively.
Mean ± SD gadolinium concentration in the brainstem (dark-gray bars), cerebellum (light-gray bars), frontal lobe white matter (white bars), parietal lobe (black bars), piriform cortex (diagonal-striped bars), and thalamus (vertical-striped bars) of 8 dogs (group 1) after IV administration of various doses of gadodiamide 3 to 7 days earlier. Notice the increase in gadolinium deposition in the brainstem and cerebellum, compared with deposition in the other regions. The relatively higher deposition for the 0.006-mmol/kg dose may have been a spurious finding.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.949
Gadolinium concentration (ng of gadolinium/g of brain tissue) in tissues obtained from each of 6 brain regions for dogs given various doses of gadodiamide IV 3 to 7 days (group 1) or appoximately 9 hours (group 2) prior to euthanasia.
| Group | Dose range (mmol/kg) | Value | Brainstem | Cerebellum | Frontal lobe | Parietal lobe | Piriform cortex | Thalamus |
|---|---|---|---|---|---|---|---|---|
| 1 (n = 8) | 0.006–0.1 | Median | 20.36 | 48.03 | 17.15 | 36.01 | 36.23 | 24.19 |
| Mean | 23.14 | 62.28 | 18.37 | 39.18 | 40.65 | 28.21 | ||
| Range | 4.19–54.53 | 10.92–162.50 | 1.71–36.18 | 7.88–79.67 | 15.47–83.12 | 7.84–51.9 | ||
| 2 (n = 5) | 0.25–0.1 | Median | 927.6 | 363.6 | 100.5 | 343.2 | 335.2 | 174.1 |
| Mean | 916.52 | 397.4 | 95.98 | 513.72 | 433.98 | 163.2 | ||
| Range | 550.0–1,216.4 | 220.80–658.90 | 68.60–112.60 | 290.70–910.30 | 289.90–839.00 | 67.30–222.60 |
Analysis of group 1 and 2 dogs
Gadolinium tissue concentration in formalin-fixed tissues was significantly (P < 0.001) affected by the group-by-region interaction in the analysis that combined group 1 and 2 dogs (Table 2). Mean gadolinium concentrations for group 1 dogs were significantly lower than concentrations for group 2 dogs for the brainstem (P < 0.001), cerebellum (P = 0.025), parietal lobe (P < 0.001), and piriform cortex (P = 0.002). There were no significant differences in mean gadolinium concentrations between groups for the frontal lobe white matter (P = 0.999) or thalamus (P = 0.975). There were no significant (P > 0.999) differences in mean concentrations among regions of the brain in group 1 dogs.
Mean gadolinium concentration (ng of gadolinium/g of brain tissue) in 6 regions of the brain of dogs of groups 1 (n = 8) and 2 (5) after IV administration of various doses of gadodiamide.
| Brain region | Group | Concentration |
|---|---|---|
| Brainstem | 1 | 25.1 ± 45.71 |
| 2 | 744.0 ± 52.83 | |
| Cerebellum | 1 | 64.2 ± 45.71 |
| 2 | 311.0 ± 52.83 | |
| Frontal lobe white matter | 1 | 20.3 ± 45.71 |
| 2 | 59.3 ± 52.83 | |
| Parietal lobe | 1 | 41.1 ± 45.71 |
| 2 | 407.9 ± 52.83 | |
| Piriform cortex | 1 | 42.6 ± 45.71 |
| 2 | 341.4 ± 52.83 | |
| Thalamus | 1 | 30.1 ± 45.71 |
| 2 | 115.5 ± 52.83 |
Values reported are least squares mean ± SE. Least squares means were derived from a linear mixed model, with dose, group, region, and the group-by-region and group-by-dose interactions as fixed effects and dog as a random effect.
In group 2 dogs, the gadolinium concentration was significantly (P < 0.001) greater in the brainstem than in all the other brain regions, and the gadolinium concentration in the cerebellum was significantly (P = 0.019) greater than in the frontal lobe white matter but did not differ significantly from the concentration in the parietal lobe (P = 0.942), piriform cortex (P = 1.000), or thalamus (P = 0.150). Gadolinium concentration in the parietal lobe of group 2 dogs was significantly greater than in the frontal lobe white matter (P < 0.001) and thalamus (P = 0.003) but did not differ significantly (P = 0.996) from the concentration in the piriform cortex. Gadolinium concentration in the piriform cortex of group 2 dogs was significantly greater than in the frontal lobe white matter (P = 0.004) but did not differ significantly (P = 0.051) from the concentration in the thalamus.
Irrespective of group, concentration in the frontal lobe white matter and thalamus did not differ significantly (P = 0.999). There was a significant (P < 0.001) group-by-dose interaction for the analysis of the combined group 1 and 2 dogs. Regression plots of tissue concentration by dose for group 1 and 2 dogs were created (Figure 2).
Regression plots of gadolinium concentration for the brainstem (black circles and solid line), cerebellum (black triangles and dotted line), frontal lobe white matter (black squares and dotted line), parietal lobe (white circles and dashed line), piriform cortex (black diamonds and dashed line), and thalamus (white squares and dashed line) of the 8 dogs of group 1 after IV administration of various doses of gadodiamide. Gadolinium concentrations were determined with ICP-MS. Within a dose, symbols represent results for tissues obtained from 1 dog. Regression lines are the gadolinium concentration predicted by use of a linear mixed model with dose, region, and the dose-by-region interaction as fixed effects and dog as a random effect.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.949
Analysis of group 1 dogs alone
Because of the significant interactions with group, data for group 1 and 2 dogs were analyzed separately. For group 1 dogs, gadolinium concentration in formalin-fixed tissue was significantly (P < 0.001) affected by the dose-by-region interaction. Among the brain regions evaluated, the concentration in the cerebellum was significantly (P < 0.001) different (ie, greater concentration) than in the other regions. Regression plots of gadolinium concentration for tissues of each of the brain regions on the basis of dose were created (Figure 3).
Regression plots of gadolinium concentration for brain tissues obtained from 6 regions of dogs of group 1 (circles and solid line; n = 8) and group 2 (triangles and dotted line; 5) after IV administration of various doses of gadodiamide 3 to 7 days (group 1) or appoximately 9 hours (group 2) earlier. Gadolinium concentrations were determined with ICP-MS. Within a dose, each symbol represents results for a brain region of 1 dog. Regression lines are the gadolinium concentration predicted by use of a linear mixed model with dose, group, region, and the group-by-region and group-by-dose interactions as fixed effects and dog as a random effect.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.949
The brains of dogs administered a single dose contained a mean of 0.028% of the total dose (group 1, 0.02%; group 2, 0.04%) across the 6 regions of brain evaluated, which likely represented only a small portion of the total gadolinium accumulation in the brain for each dog. There was significant affinity for the cerebellum in dogs of group 1 and affinity for the brainstem and cerebellum in dogs of group 2. Furthermore, retention of gadolinium in the cerebellum ranged from 10.92 to 162.5 ng of gadolinium/g of brain tissue in dogs of group 1 and from 220.8 to 658.9 ng of gadolinium/g of brain tissue in dogs of group 2.
Analysis of group 2 dogs alone
In group 2 dogs, gadolinium concentration in formalin-fixed tissue was significantly (P < 0.001) affected by region (Table 3). Gadolinium concentration was significantly greater in the brainstem than in the cerebellum (P = 0.005), frontal lobe white matter (P < 0.001), parietal lobe (P = 0.043), piriform cortex (P = 0.009), and thalamus (P < 0.001). Gadolinium concentration in the cerebellum did not differ significantly from the concentration in the frontal lobe white matter (P = 0.210), parietal lobe (P = 0.941), piriform cortex (P = 1.000), or thalamus (P = 0.465). Gadolinium concentration in the frontal lobe white matter was significantly (P = 0.031) lower than the concentration in the parietal lobe but did not differ significantly from the concentration in the piriform cortex (P = 0.127) or thalamus (P = 0.995). Gadolinium concentration in the parietal lobe did not differ significantly from the concentration in the piriform cortex (P = 0.989) or thalamus (P = 0.104). Gadolinium concentration in the piriform cortex did not differ significantly (P = 0.314) from the concentration in the thalamus. Gadolinium concentrations were also significantly (P < 0.001) affected by dose; however, in contrast to group 1 dogs, there was not a significant (P = 0.123) dose-by-region effect. Regression plots of gadolinium concentration by dose administered to group 1 and group 2 dogs were created (Figure 3).
Gadolinium concentration (ng of gadolinium/g of brain tissue) in 6 regions of the brain of the 5 dogs of group 2 after IV administration of various doses of gadodiamide and examination of formalin-fixed brain tissues.
| Brain region | Concentration |
|---|---|
| Brainstem | 764.7 ± 81.67 |
| Cerebellum | 331.7 ± 81.67 |
| Frontal lobe white matter | 80.0 ± 81.67 |
| Parietal lobe | 428.6 ± 81.67 |
| Piriform cortex | 362.1 ± 81.67 |
| Thalamus | 136.2 ± 81.67 |
Values reported are least squares mean ± SE. Least squares means were derived from a linear mixed model with dose and region as fixed effects and dog as a random effect.
Analysis of fresh-frozen versus formalin-fixed tissue
Brain samples of all group 2 dogs were stored in formalin for 69 days before ICP-MS, whereas fresh-frozen tissues were stored for 32 days before ICP-MS. Gadolinium concentrations did not differ significantly (P = 0.430) between formalin-fixed and fresh-frozen tissues of group 2 dogs. There was a significant (P < 0.001) effect on concentration by brain region, irrespective of tissue fixation status or dose (Table 4). Gadolinium concentration in the brainstem was significantly greater than in the frontal lobe white matter (P < 0.001), piriform cortex (P = 0.007), and thalamus (P < 0.001) but did not differ significantly from the concentration in the cerebellum (P = 0.068) or parietal lobe (P = 0.266). Gadolinium concentration in the cerebellum was significantly greater than in the frontal lobe white matter (P = 0.016) but did not differ significantly from the concentration in the parietal lobe (P = 0.987), piriform cortex (P = 0.955), or thalamus (P = 0.097). Gadolinium concentration in the frontal lobe white matter was significantly (P = 0.003) lower than in the parietal lobe but did not differ significantly from the concentration in the piriform cortex (P = 0.144) or thalamus (P = 0.982). Gadolinium concentration in the piriform cortex did not differ significantly from the concentration in the parietal lobe (P = 0.673) or thalamus (P = 0.474). Gadolinium concentration in the parietal lobe was significantly (P = 0.019) greater than the concentration in the thalamus. Gadolinium concentrations were also significantly (P = 0.003) affected by dose. Regression plots of gadolinium concentration by dose, as determined by use of models that included dose and region for group 2 dogs, were created for the 2 tissue-type analyses (Figure 4).
Regression plots of gadolinium concentration for brain tissues obtained from 6 regions of the 5 dogs of group 2 after IV administration of gadodiamide at various doses. Regression lines are the gadolinium concentration predicted by use of a linear mixed model with dose and region as fixed effects and dog as a random effect (A) and by use of a linear mixed model with tissue type, dose, and region as fixed effects and dog as a random effect (B). Gadolinium concentrations were determined with ICP-MS. Within a dose, each symbol represents results for a brain region of 1 dog.
Citation: American Journal of Veterinary Research 79, 9; 10.2460/ajvr.79.9.949
Gadolinium concentration (ng of gadolinium/g of brain tissue) in 6 regions of the brain of the 5 dogs of group 2 after IV administration of various doses of gadodiamide and examination of fresh-frozen and formalin-fixed brain tissues.
| Brain region | Concentration |
|---|---|
| Brainstem | 618.4 ± 74.45 |
| Cerebellum | 379.4 ± 74.45 |
| Frontal lobe white matter | 95.0 ± 74.45 |
| Parietal lobe | 433.5 ± 74.45 |
| Piriform cortex | 306.4 ± 74.45 |
| Thalamus | 153.4 ± 74.45 |
Values reported are least squares mean ± SE. Least squares means were derived from a linear mixed model with tissue (fresh-frozen or formalin-fixed tissues), dose, and region as fixed effects and dog as a random effect.
Localization of gadolinium in brain tissues and assessment of histologic changes
Gadolinium was not detected during SEM evaluation in any samples from the dogs of group 1. Six samples from group 2 dogs (cerebellum and parietal lobe of dogs that received gadodiamide at 0.1 [2 dogs] or 0.05 [1 dog] mmol/kg) were evaluated with TEM. There were several angular, irregularly shaped structures with a mean length of 1,009 nm (range, 683 to 1,360 nm) throughout the neuronal interstitium of the examined tissue. These structures had an L-series K-edge value of 2.111, which is specific to elemental gadolinium, and corresponded to approximately 0.19% of the sample weight. Other elements found in the samples included carbon, oxygen, sodium, magnesium, silicon, sulfur, chlorine, cobalt, nickel, copper, and lead, which were attributed to physiologic factors or sample processing.
Discussion
In the study reported here, gadolinium retention in brain tissue of healthy dogs after a single IV administration of gadodiamide (a linear nonionic GBCA) at various doses (range, 0.006 to 0.1 mmol/kg) was evaluated by use of ICP-MS, which is the evaluation test of choice for gadolinium quantification in bone, skin, and brain.3,21,23,31,32 Gadolinium was detected after administration of the lowest fractional dose (0.006 mmol/kg). Neuronal retention of gadolinium in humans has been repeatedly documented, especially for linear agents. The authors are aware of no previous studies on deceased dogs or humans that were conducted to evaluate neural retention of gadolinium after IV administration of gadodiamide to healthy subjects that had no evidence of serious systemic disease, brain tumors, treatment with intracranial radiation, or multiorgan failure. Access to healthy human brain tissue of those who received a GBCA is limited.3,8,21,23
The GBCAs are generally classified on the basis of their structure (linear vs macrocyclic) and charge (nonionic vs ionic).7 The chelated ligand is not fully closed around the gadolinium ion in the open-chain (ie, linear) classification, whereas it fully encloses the gadolinium ion in the macrocyclic configuration, which is a configuration that contributes to greater thermodynamic, kinetic, and physiologic stability.7,8 Although nonionic iodinated contrast agents are generally preferred because of their safety, ionic GBCAs are more stable.7,8 When single or multiple doses of nonionic macrocyclic or ionic linear agents were compared, higher deposition was found with the linear agents, although deposition occurred with both linear and macrocyclic agents (gadobutrol, gadoteridol, gadoxetate, and gadobenate).23 Multiple administrations of a nonionic linear agent (gadodiamide) also result in brain deposition without significant differences among locations of deposition.3 Differences in signal intensity persistence within the brain parenchyma of subjects receiving nonionic and ionic linear agents exist, with the linear agents having greater persistence.11 Administration of at least 20 to 41 doses of macrocyclic agents (gadoterate and gadobutrol) has not resulted in persistent parenchymal enhancement.33
For reasons that are poorly understood, the degree of gadolinium retention is affected by dose and region of the brain, with certain regions, including the cerebellum, parietal lobe, and brainstem, having greater retention. A study3 of humans found that the dentate nucleus had the highest dose-dependent T1-weighted change in signal intensity, which correlates with results of the study reported here in which we found greater dose-dependent deposition in the cerebellum of dogs. Retention of gadolinium in the cerebellum ranged from 10.92 to 162.5 ng of gadolinium/g of brain tissue in dogs of group 1 and from 220.8 to 658.9 ng of gadolinium/g of brain tissue in dogs of group 2. This is similar to results reported for people undergoing clinically indicated MRI; those results include a median concentration of 6.6 μg of gadolinium/g of brain tissue (range, 0.1 to 58.8 μg of gadolinium/g of brain tissue),3 a range of 0.1 to 2.1 μg of gadolinium/g of brain tissue,21 and a range of 0.003 to 0.107 μg of gadolinium/g of brain tissue.23
Results of the present study support the reported relative affinity of gadolinium for the cerebellum regardless of patient health status.15 When compared with results for deceased humans, the pattern of retention in brain tissue for dogs of group 1 mimicked those described in people, specifically with regard to dose-dependent effects in the thalamus, white matter of the frontal lobe, and cerebellum.3 Furthermore, the present study confirmed that gadolinium retention was lower in dogs for which there was a greater interval between gadodiamide administration and tissue collection. Finally, the dose effect was affected by the interval between gadodiamide administration and tissue collection, with the dose effect being greater for group 1 dogs (9 hours between administration and tissue collection) than for group 2 dogs (3 to 7 days between administration and tissue collection). Expanding our understanding of the characteristics and patterns of retention in the brain for various doses and classes of GBCAs in healthy subjects without disruption of the blood-brain barrier may help elucidate the incompletely understood pharmacokinetics and dynamics of a drug that has been presumed to not result in deposits in intracranial tissues.
In humans and other animals, pharmacokinetics of GBCAs is similar to the pharmacokinetics of iodinated radiographic contrast agents.2 Gadolinium has a mean ± SD extracellular distribution similar to that of extracellular water (approx 200 ± 61 mL/kg), with rapid, unchanged renal elimination by way of glomerular filtration (> 90% recovery within 3 hours and 95.5 ± 5% recovery within 24 hours).34 Gadolinium has a short half-life (1 to 2 hours), and the remaining gadolinium is cleared by hepatic clearance in the bile.2,34–37 Renal and plasma clearance rates (1.7 and 1.8 mL/min/kg, respectively) are nearly identical.34 Gadodiamide conforms to a 2-compartmental (extracellular) model, with the mean ± SD for the mean distribution and elimination half-lives of 3.4 ± 2.7 minutes and 77.8 ± 16 minutes, respectively. Detectable biotransformation or dissociation of GBCAs has not been reported.34
Since their inception, it has been thought that GBCAs do not cross the intact blood-brain barrier, and it was presumed that they did not accumulate in the brain or in brain lesions in which there was a normal blood-brain barrier (eg, cysts or mature postoperative scar tissue).34 Findings of the study reported here and studies of humans3,31,32 suggest a more complicated pharmacokinetic behavior than has previously been believed. The physiologic reasons behind differences in tissue distribution (and possibly redistribution) of gadolinium between the 2 groups of dogs in the present study are not understood. It has been proposed that a washout period be imposed between subsequent administrations of GBCAs because signal intensity in people who receive ionic linear agents decreases over time if no additional GBCAs are administered.38 It is possible that a washout period could be a contributing factor in the patterns and amounts of retained elemental gadolinium detected in the dogs of the study reported here.
Gadolinium retention was not detected in fresh-frozen tissues of the control dog, as determined by use of ICP-MS. However, formalin-fixed samples of the control dog had trace amounts of gadolinium in the brainstem, cerebellum, parietal lobe, piriform cortex, and thalamus. This is in agreement with results of gadolinium contamination in control patients, for which ICP-MS detected small amounts of gadolinium deposition (up to 0.2 ng of gadolinium/g of tissue) in numerous brain regions in human patients who had not been given GBCAs.23 Similar to the results for humans, the trace amounts detected in the control dog were orders of magnitude lower than the concentrations in the exposed dogs. This has historically been attributed to minute environmental exposures.23 Given our inability to confirm these concentrations in the fresh-frozen tissues from the control dog, we attributed these trace amounts to contamination during sample excision, preparation, fixation, or processing or residue on the machines or surfaces at any step of processing or any point during ICP-MS.
When results for fresh-frozen tissues were compared with results for formalin-fixed tissues of group 2 dogs, no significant difference between gadolinium retention was detected between tissues. There was also no difference between retention and dose, but regional differences were noted, with the cerebellum and brainstem having the highest concentrations. We concluded that for short-term (< 70 days) storage, there were no significant effects as a result of leaching of gadolinium into the formalin fixative solution. However, effects of a longer duration (months to years) of sample storage in formalin solution on quantification of gadolinium retention in tissues remain unknown.
The mechanism by which a clinically normal brain retains gadolinium is poorly understood. Brain-wide paravascular or glymphatic exchange of CSF and interstitial fluid for clearance of solutes and waste has been described in mice.39,40 In 1 study,39 gadolinium allowed for the identification of waste influx at key anatomic influx nodes, specifically the pituitary and pineal gland recesses. In addition, kinetic parameters that characterize influx and clearance routes of paramagnetic contrast agents have been defined, and glymphatic exchange of CSF and interstitial fluid for solute clearance was described.39 Secretion and reabsorption of CSF is not limited to traditional CSF secretion modeling and a historical understanding of kinematics wherein there is only antegrade flow of CSF.40
In mice, a large portion of subarachnoid CSF recirculates throughout the brain parenchyma along paravascular spaces and is exchanged with the interstitial fluid along such routes and through the interstitium by means of transglial water movement through astrocytic aquaporin-4 water channels, which facilitates the clearance of interstitial solutes.40 Retrograde paravascular influx of paramagnetic contrast agents from the subarachnoid space into the brain parenchyma occurs rapidly by way of Virchow-Robin spaces along the para-arterial channels.40 In the authors’ opinion, these regions and pathways provide a conduit for solute uptake from the brain as well as potentially a mechanism for retention within the brain parenchyma. Although it has not been proven, clearance pathways identified by use of gadolinium-enhanced MRI may provide evidence on the potential pathways of parenchymal exposure and retention. Furthermore, although glymphatic flow in unconscious or anesthetized dogs is not well characterized, it is plausible that anesthesia could speed the rate of glymphatic flow that contributes to differences of deposition in conscious versus anesthetized animals.41
Important factors that may contribute to gadolinium retention in the brain are the physical state of gadolinium at the time of parenchymal exposure, physiologic mechanisms or events that drive retention (which may be unique to each patient and disease state), and overall sequestration, deposition, and neuroanatomic affinity of elemental and chelated gadolinium. These factors remain incompletely understood.3 Furthermore, the physical state of the element cannot be speciated in tissue by use of currently available methods.3,21,23 Chelates are historically considered stable pharmaceutical entities. Results of the study reported here and other studies3,23 support that they may not be stable entities. Gadodiamide does not bind to human serum proteins in vitro, and GBCAs do not undergo biotransformation (and there is no evidence of metabolism) in urine or feces.42,43 Radiocarbon tracing of gadodiamide in mice revealed that for a given organ compartment, elimination of the compound depends on the specific organ, with identical clearance of gadolinium and the ligand (when measured separately).42,44 Thus, the conclusion that disassociation does not take place is erroneously based on a premise that lacks more specific speciation of the excreted components (the ligand and the ion). Furthermore, it is highly plausible that in vitro findings cannot be extrapolated to in vivo pharmacokinetics for these agents.44,45
The present study had a number of limitations. Sample numbers were small, and the dogs evaluated were juveniles. The study population was limited to the number of available purpose-bred dogs intended for use in terminal surgical laboratories. This also limited the dosing regimen, the ability to evaluate total brain gadolinium retention, and the number of sites investigated (precluding investigation of specific neuronal bodies of each of the cerebral lobes and retention of gadolinium in other viscera). We were also limited to use of a single agent (gadodiamide) such that the findings reported here may not be applicable to macrocyclic GBCAs. These limitations are challenging and typically preclude prospective studies of this nature in veterinary medicine. A larger subject pool likely would have reduced variability of sample measurements. The authors chose to evaluate gadolinium accumulation in brain tissue after a dose at or less than the standard dose frequently administered to veterinary patients assessed at our institution. It was believed that 1 control dog would be sufficient for verification of negative natural exposure and validation of quantification methods. Tissues from the control dog were not evaluated with SEM and TEM for comparison with gadodiamide-injected dogs. Although results of prescreening tests of hepatic and renal function of all dogs were within reference ranges, calculation of estimated glomerular filtration rate and nuclear scintigraphy or CT determination of glomerular filtration rate were not performed. A relevant equation for estimation of glomerular filtration rate across breeds or species is not currently available. We had no reason to suspect that the dogs of the study population had underlying renal insufficiency. Finally, we were unable to discern the physical chelation state of the neuronal gadolinium deposited in these dogs. This is a contemporary technological limitation that is not specific to our institution. Speciation of neuronal gadolinium deposits in formalin-fixed tissue is currently unavailable.3,42
In the study reported here, a single dose of gadodiamide administered IV was associated with dose-dependent retention in the neuropil unrelated to renal and hepatobiliary function in young healthy dogs. Affinity was especially evident in the cerebellum and brainstem. Evaluation of brain tissue that had been stored in formalin for < 70 days did not reveal significant effects on gadolinium quantification, and leaching of gadolinium into the formalin fixative solution was presumed to be minimal during short-term storage. Although we determined that intracranial gadolinium was present, we were unable to determine the exact cell type or location. Furthermore, we could not determine the histologic phenotype of the deposited gadolinium. Fractional doses resulted in gadolinium retention in the brain of healthy uncompromised canine subjects. Long-term effects of repeated doses of GBCAs in companion animals at or below standard doses, such as those used in the present study, remain unknown; however, in the authors’ opinion, they are unlikely to be of consequence given the typical life span of dogs. The authors are not aware of any reports that provide a description of the biological or clinical importance of this deposition; however, there is still concern that there may be delayed effects.8 Of the 400 intracranial diseases described in people, approximately 300 have been identified in dogs, which make them a viable species for investigation of intracranial diseases of humans.46 Despite the possibility that access to dogs for use in terminal veterinary teaching laboratories and research experiments is likely to become more limited for ethical reasons, such dogs can be used in in vivo translational investigations of retention profiles of other GBCAs.
Furthermore, results reported for the present study may aid clinicians when making decisions regarding the choice of a GBCA (ie, linear nonionic, linear ionic, macrocyclic nonionic, and macrocyclic ionic). Future investigations of large cohorts conducted after patients have received these agents and their outcomes are needed to determine whether gadolinium retention poses a safety risk or long-term health consequence for humans and other animals. Results of the present study may serve as an impetus for evaluating other classes of GBCAs, investigating mechanisms by which GBCAs cross the blood-brain barrier, and exploring alternative means of increasing the sensitivity and specificity of MRI.
Acknowledgments
This manuscript represents a portion of a thesis submitted by Dr. Lee to the graduate school of the Department of Clinical Sciences, College of Veterinary Medicine, Mississippi State University, as partial fulfillment of the requirements for a Master of Veterinary Science degree.
Funded in part by a Resident Research Award Grant from the CT/MR society of the American College of Veterinary Radiology and by Dr. Gambino thorough the Department of Clinical Sciences, College of Veterinary Medicine, Mississippi State University.
The authors declare that there were no conflicts of interest.
Presented at the American College of Veterinary Radiology Annual Scientific Conference, Orlando, Fla, October 2016.
The authors thank Dr. Kristen Fizzano for assistance with anesthesia, Dr. Bridgett Willeford for assistance with animal care, Gary Sorrels and Molly Nicholson for assistance with MRI procedures, Emerald Barrett for assistance with data acquisition and analysis, and Dr. Mary E. Milewski for technical assistance.
ABBREVIATIONS
| GBCA | Gadolinium-based chelated contrast agent |
| ICP-MS | Inductively coupled mass spectrometry |
| SEM | Scanning electron microscopy |
| TEM | Transmission electron microscopy |
Footnotes
CellDyn 3700, Abbott Diagnostics, Santa Clara, Calif.
Vet Axcell, Alfa Wasserman Inc, West Caldwell, NJ.
HTW-Ag test, ProtaTek Reference Laboratory, Mesa, Ariz.
ProtaTek Reference Laboratory, Mesa, Ariz.
Diagnostic Center for Population and Animal Health, College of Veterinary Medicine, Michigan State University, Lansing, Mich.
Omniscan, 0.5 mmol/mL, General Electric Healthcare, Princeton, NJ.
Baxter Healthcare, Deerfield, Ill.
Beuthanasia-D Special, Schering-Plough Animal Health Corp, Kenilworth, NJ.
Kimwipe, Kimberly-Clark Corp, Neenah, Wis.
Lindberg/Blue M G01305A, VWR, Radnor, Pa.
Agilent 7500ce inductively coupled plasma-mass spectrometer, Agilent Technologies, Santa Clara, Calif.
Inorganic Ventures, Christiansburg, Va.
Ultracut E, Reichert-Jung Inc, Depew, NY.
JEM-1230, JEOL USA, Peabody, Mass.
JEM-2100, JEOL USA, Peabody, Mass.
AZtecTEM, Oxford Instruments, Oxford, England.
Zeiss EVO-50 VP SEM, Zeiss, Jena, Germany.
Quantax EDS, Bruker Corp, Billerica, Mass.
PROC MIXED, SAS for Windows, version 9.4, SAS Institute Inc, Cary, NC.
Simulate, SAS for Windows, version 9.4, SAS Institute Inc, Cary, NC.
References
1. Kuriashkin IV, Losonsky JM. Contrast enhancement in magnetic resonance imaging using intravenous paramagnetic contrast media: a review. Vet Radiol Ultrasound 2000;41:4–7.
2. Brady TJ, Reimer P. Contrast agents in whole body magnetic resonance: an overview. In: Young IR, ed. Methods in biomedical magnetic resonance imaging and spectroscopy. Ann Arbor, Mich: Wiley, 2000;693–698.
3. McDonald RJ, McDonald JS, Kallmes DF, et al. Intracranial gadolinium deposition after contrast-enhanced MR imaging. Radiology 2015;275:772–782.
4. Robertson I. Optimal magnetic resonance imaging of the brain. Vet Radiol Ultrasound 2011;52:S15–S22.
5. Niendorf H, Brasch R. Gd-DTPA tolerance and clinical safety. In: Brasch R, ed. MRI contrast enhancement in the central nervous system: a case study approach. New York: Raven Press, 1993;11–21.
6. Caravan P, Ellison JJ, McMurry TJ, et al. Gadolinium(III) chelates as MRI contrast agents: structure, dynamics, and applications. Chem Rev 1999;99:2293–2352.
7. Ramalho J, Semelka RC, Ramalho M, et al. Gadolinium-based contrast agent accumulation and toxicity: an update. AJNR Am J Neuroradiol 2016;37:1192–1198.
8. Gulani V, Calamante F, Shellock FG, et al. Gadolinium deposition in the brain: summary of evidence and recommendations. Lancet Neurol 2017;16:564–570.
9. Robert P, Lehericy S, Grand S, et al. T1-weighted hypersignal in the deep cerebellar nuclei after repeated administrations of gadolinium-based contrast agents in healthy rats: difference between linear and macrocyclic agents. Invest Radiol 2015;50:473–480.
10. Radbruch A, Weberling LD, Kieslich PJ, et al. Gadolinium retention in the dentate nucleus and globus pallidus is dependent on the class of contrast agent. Radiology 2015;275:783–791.
11. Ramalho J, Castillo M, AlObaidy M, et al. High signal intensity in globus pallidus and dentate nucleus on unenhanced T1-weighted MR images: evaluation of two linear gadolinium-based contrast agents. Radiology 2015;276:836–844.
12. Ray DE, Holton JL, Nolan CC, et al. Neurotoxic potential of gadodiamide after injection into the lateral cerebral ventricle of rats. AJNR Am J Neuroradiol 1998;19:1455–1462.
13. Roman-Goldstein SM, Barnett PA, McCormick CI, et al. Effects of gadopentetate dimeglumine administration after osmotic blood-brain barrier disruption: toxicity and MR imaging findings. AJNR Am J Neuroradiol 1991;12:885–890.
14. Kanal E, Tweedle MF. Residual or retained gadolinium: practical implications for radiologists and our patients. Radiology 2015;275:630–634.
15. Kanda T, Ishii K, Kawaguchi H, et al. High signal intensity in the dentate nucleus and globus pallidus on unenhanced T1-weighted MR images: relationship with increasing cumulative dose of a gadolinium-based contrast material. Radiology 2014;270:834–841.
16. Errante Y, Cirimele V, Mallio CA, et al. Progressive increase of T1 signal intensity of the dentate nucleus on unenhanced magnetic resonance images is associated with cumulative doses of intravenously administered gadodiamide in patients with normal renal function, suggesting dechelation. Invest Radiol 2014;49:685–690.
17. Quattrocchi CC, Mallio CA, Errante Y, et al. Gadodiamide and dentate nucleus T1 hyperintensity in patients with meningioma evaluated by multiple follow-up contrast-enhanced magnetic resonance examinations with no systemic interval therapy. Invest Radiol 2015;50:470–472.
18. Kanda T, Osawa M, Oba H, et al. High signal intensity in dentate nucleus on unenhanced T1-weighted MR images: association with linear versus macrocyclic gadolinium chelate administration. Radiology 2015;275:803–809.
19. Roberts DR, Holden KR. Progressive increase of T1 signal intensity in the dentate nucleus and globus pallidus on unenhanced T1-weighted MR images in the pediatric brain exposed to multiple doses of gadolinium contrast. Brain Dev 2016;38:331–336.
20. Miller JH, Hu HH, Pokorney A, et al. MRI brain signal intensity changes of a child during the course of 35 gadolinium contrast examinations. Pediatrics 2015;136:e1637–e1640.
21. Kanda T, Fukusato T, Matsuda M, et al. Gadolinium-based contrast agent accumulates in the brain even in subjects without severe renal dysfunction: evaluation of autopsy brain specimens with inductively coupled plasma mass spectroscopy. Radiology 2015;276:228–232.
22. Adin ME, Kleinberg L, Vaidya D, et al. Hyperintense dentate nuclei on T1-weighted MRI: relation to repeat gadolinium administration. AJNR Am J Neuroradiol 2015;36:1859–1865.
23. Murata N, Gonzalez-Cuyar LF, Murata K, et al. Macrocyclic and other non-group 1 gadolinium contrast agents deposit low levels of gadolinium in brain and bone tissue: preliminary results from 9 patients with normal renal function. Invest Radiol 2016;51:447–453.
24. Malayeri AA, Brooks KM, Bryant LH, et al. National Institutes of Health perspective on reports of gadolinium deposition in the brain. J Am Coll Radiol 2016;13:237–241.
25. Kasahara S, Miki Y, Kanagaki M, et al. Hyperintense dentate nucleus on unenhanced T1-weighted MR images is associated with a history of brain irradiation. Radiology 2011;258:222–228.
26. US FDA. FDA evaluating the risk of brain deposits with repeated use of gadolinium-based contrast agents for magnetic resonance imaging (MRI). Safety announcement. Silver Spring, Md: US Department of Health and Human Services, 2015.
27. European Medicines Agency. EMA's final opinion confirms restrictions on use of linear gadolinium agents in body scans. London: European Medicines Agency, 2017.
28. US FDA. FDA drug safety communication: FDA identifies no harmful effects to date with brain retention of gadolinium-based contrast agents for MRIs; review to continue. Silver Spring, Md: US Department of Health and Human Services, 2017.
29. Dickinson PJ, LeCouteur RA, Higgins RJ, et al. Canine spontaneous glioma: a translational model system for convection-enhanced delivery. Neuro Oncol 2010;12:928–940.
30. Potschka H, Fischer A, von Rüden E-L, et al. Canine epilepsy as a translational model? Epilepsia 2013;54:571–579.
31. White GW, Gibby WA, Tweedle MF. Comparison of Gd(DTPA-BMA) (Omniscan) versus Gd(HP-DO3A) (ProHance) relative to gadolinium retention in human bone tissue by inductively coupled plasma mass spectroscopy. Invest Radiol 2006;41:272–278.
32. Darrah TH, Prutsman-Pfeiffer JJ, Poreda RJ, et al. Incorporation of excess gadolinium into human bone from medical contrast agents. Metallomics 2009;1:479–488.
33. Radbruch A, Haase R, Kieslich PJ, et al. No signal intensity increase in the dentate nucleus on unenhanced T1-weighted MR images after more than 20 serial injections of macrocyclic gadolinium-based contrast agents. Radiology 2017;282:699–707.
34. Omniscan [package insert]. Oslo: GE Healthcare, 2010.
35. Aime S, Caravan P. Biodistribution of gadolinium-based contrast agents, including gadolinium deposition. J Magn Reson Imaging 2009;30:1259–1267.
36. Weinmann HJ, Laniado M, Mutzel W. Pharmacokinetics of GdDTPA/dimeglumine after intravenous injection into healthy volunteers. Physiol Chem Phys Med NMR 1984;16:167–172.
37. Oksendal AN, Hals P-A. Biodistribution and toxicity of MR imaging contrast media. J Magn Reson Imaging 1993;3:157–165.
38. Radbruch A, Weberling LD, Kieslich PJ, et al. Intraindividual analysis of signal intensity changes in the dentate nucleus after consecutive serial applications of linear and macrocyclic gadolinium-based contrast agents. Invest Radiol 2016;51:683–690.
39. Iliff JJ, Lee H, Yu M, et al. Brain-wide pathway for waste clearance captured by contrast-enhanced MRI. J Clin Invest 2013;123:1299–1309.
40. Iliff JJ, Wang M, Liao Y, et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid beta. Sci Transl Med 2012;4:147ra111.
41. Mendelsohn AR, Larrick JW. Sleep facilitates clearance of metabolites from the brain: glymphatic function in aging and neurodegenerative diseases. Rejuvenation Res 2013;16:518–523.
42. Telgmann L, Sperling M, Karst U. Determination of gadolinium-based MRI contrast agents in biological and environmental samples: a review. Anal Chim Acta 2013;764:1–16.
43. Arbughi T, Bertani F, Celeste R, et al. High-performance liquid chromatographic determination of the magnetic resonance imaging contrast agent gadobenate ion in plasma, urine, faeces, bile and tissues. J Chromatogr B Biomed Sci Appl 1998;713:415–426.
44. Kindberg GM, Uran S, Friisk G, et al. The fate of Gd and chelate following intravenous injection of gadodiamide in rats. Eur Radiol 2010;20:1636–1643.
45. Normann PT, Hals PA. In vivo stability and excretion of gadodiamide (GdDTPA-BMA), a hydrophilic gadolinium complex used as a contrast enhancing agent for magnetic resonance imaging. Eur J Drug Metab Pharmacokinet 1995;20:307–313.
46. Mata M, Gottschalk S. Man's best friend: utilizing naturally occurring tumors in dogs to improve chimeric antigen receptor T-cell therapy for human cancers. Mol Ther 2016;24:1511–1512.
