How to stop using gadolinium chelates for magnetic resonance imaging: clinical-translational experiences with ferumoxytol

Abstract

Gadolinium chelates have been used as standard contrast agents for clinical MRI for several decades. However, several investigators recently reported that rare Earth metals such as gadolinium are deposited in the brain for months or years. This is particularly concerning for children, whose developing brain is more vulnerable to exogenous toxins compared to adults. Therefore, a search is under way for alternative MR imaging biomarkers. The United States Food and Drug Administration (FDA)-approved iron supplement ferumoxytol can solve this unmet clinical need: ferumoxytol consists of iron oxide nanoparticles that can be detected with MRI and provide significant T1- and T2-signal enhancement of vessels and soft tissues. Several investigators including our research group have started to use ferumoxytol off-label as a new contrast agent for MRI. This article reviews the existing literature on the biodistribution of ferumoxytol in children and compares the diagnostic accuracy of ferumoxytol- and gadolinium-chelate-enhanced MRI. Iron oxide nanoparticles represent a promising new class of contrast agents for pediatric MRI that can be metabolized and are not deposited in the brain.

Keywords: Children; Contrast agent; Ferumoxytol iron oxide nanoparticles; Magnetic resonance imaging.

Figure 1:. Ferumoxytol-MRI provides excellent vessel enhancement for MR angiographies.

Axial T1-weighted spoiled gradient recalled echo SPGR image at 10-15 minutes after intravenous infusion of ferumoxytol demonstrated a right parietal arteriovenous malformation in a 9-year-old boy. Note that for vascular imaging applications, a dose of 1-3 mg Fe/kg is usually sufficient. This patient received a dose of 3 mg Fe/kg.

Figure 2:. Comparison of tumor enhancement after intravenous administration of Gadolinium-chelate or ferumoxytol:

An osteosarcoma of the right proximal femur in 23 year old female demonstrates (a) hypointense signal on plain T1-weighted fast spin echo (FSE) image (arrow) as well as (b) hyperintense signal with regards to muscle as an internal standard on axial fat saturated T2-weighted FSE image and (c) coronal short inversion time inversion recovery (STIR) image. (d) Following intravenous administration of ferumoxytol (dose 5 mg Fe/kg), the peripheral tumor tissue demonstrates hypointense (dark) enhancement on T2-weighted FSE image, with sparing of a central area of necrosis (arrow). (e) A coronal T1-weighted scan and (f) an axial T1-weighted scan after intravenous injection of Gadolinium-chelate demonstrate corresponding peripheral enhancement of the tumor tissue (arrow), sparing of a central area of necrosis. (g) A coronal T1-weighted FSE image demonstrates hyperintense (positive) tumor enhancement, which is less intense compared to the Gadolinium-enhanced scan (e, f). (h) A delayed axial T1-weighted LAVA image about 10 minutes later demonstrates increasing tumor T1-enhancement over time (arrow), indicative of the accumulation of the nanoparticles in the tumor.

Figure 3:. Comparison of tumor enhancement after intravenous administration of Gadolinium-chelate or ferumoxytol in a 25-year-old male with desmoplastic small round cell tumour:

(a) Axial T2-weighted MR image demonstrates intermediate T2-signal of the liver and multiple T2-hyperlintense focal tumor lesions along the liver capsule and the peritoneal lining. (b) A Gadolinium-chelate enhanced axial LAVA scan demonstrates moderate contrast enhancement of both liver and focal tumor lesions. (c) Ferumoxytol-enhanced axial T2-weighted fast spin echo (FSE) image shows negative (dark) enhancement of the liver and increased tumor-to-liver contrast. (d) Axial ferumoxytol-enhanced T1-weighted LAVA image shows less enhancement of the tumor compared to the liver, thereby increasing the tumor-to-liver contrast. (e) Simultaneously acquired axial ¹⁸F-labeled fluorodeoxyglucose positron emission tomography (¹⁸F-FDG PET) images was superimposed on T1-weighted axial ferumoxytol-enhanced LAVA images and confirm the hypermetabolic (yellow and red) tumor nodules. The marked contrast enhancement of the abdominal vessels enables clear localization of tumor nodules with regards to enhancing vessels.

Figure 4:. Comparison of tumor enhancement after intravenous administration of Gadolinium-chelate or ferumoxytol in a 15-year-old male with Ewing sarcoma of the left scapula:

(a) Axial T1-weighted fast spin echo (FSE) image demonstrates hypointense intra- and extraosseous soft tissue mass in the body of the left scapula. (b) Axial T1-weighted LAVA image after intravenous administration of Gadolinium-chelate demonstrates peripheral enhancement of the tumor tissue. (c) Axial T1-weighted LAVA image after intravenous administration of ferumoxytol demonstrates marked enhancement of the central tumor necrosis (arrow), which is hyperintense compared to muscle as an internal reference standard. (d) Simultaneously acquired axial ¹⁸F-FDG PET image superimposed on T1-weighted ferumoxytol-enhanced LAVA image demonstrates the hypermetabolic (yellow) peripheral tumor tissue and central photopenic area of necrosis.

Figure 5:. Ferumoxytol-enhanced MRI can estimate presence of tumor associated macrophages in a 24-year-old male patient with Ewing sarcoma in the sacrum.

(a) Oblique coronal T2-weighted FSE image before ferumoxytol injection shows hyperintense (bright) tumor signal (arrow) compared with skeletal muscle. (b) Oblique coronal T2-weighted FSE image at 30 minutes after intravenous ferumoxytol infusion demonstrates hypointense (dark) enhancement of the hematopoietic marrow and improved outline of the hyperintense tumors (arrow). (c) Oblique coronal T2-weighted FSE image at 24 hours after ferumoxytol infusion shows hypointense (dark) tumor signal enhancement (arrows). (d) Axial T1-weighted LAVA image after intravenous ferumoxytol injection demonstrates mild T1-enhancement of the tumor (arrow), less than muscle as an internal standard tissue. This limited T1-enhancement in conjunction with strong T2-enhancement suggests predominant intracellular localization of ferumoxytol nanoparticles in tumor associated macrophages. Note that the highly cellular bone marrow does not demonstrate any T1-effect. (e) Axial ¹⁸F-fluorodeoxyglucose positron emission tomography / computed tomography (¹⁸F-FDG PET/CT) demonstrates homogenous radiotracer uptake throughout the lesion (arrow). (f) Biopsy specimen of the tumor stained with CD163 mAb demonstrated multiple CD163 positive macrophages (brown staining) in the tumor tissue (20x magnification).

Figure 6:

Ferumoxytol-labeled cell transplants can be detected in the decompression track after core decompression. (a) Coronal T2-weighted MR image of the left femur of a patient who was treated with core decompression and injection of iron-labeled cells into the decompression track. Areas of hypointense (dark) signal (arrow) are noted in the decompression track, consistent with delivery of iron-labeled cells. (b) Superimposed color-coded signal intensities show areas of iron-labeled cells (arrow) as displayed by blue color. (c) Coronal T2-weighted MR of the left femur of a patient, who was treated with core decompression and injection of unlabeled cells into the decompression track. Unlabeled cells are noted by an intermediate signal in the decompression track. (d) Superimposed color-coded signal intensities show medium ranged signal intensities (green/yellow). Signal-to-noise ratios (SNR; e) and T2_ relaxation times (f) during the first week after surgery for areas that showed iron signal compared with areas where iron-labeled cells were not delivered and unlabeled controls show significantly lower SNR (P . 0.002, n . 18; P . 0.002, n . 10; respectively) and T2 relaxation times (P . 0.007, n . 9; P . 0.02, n . 6; respectively). SNR (g) and T2_ relaxation times (h) at 4–7 weeks reveal no significant differences between the groups, suggesting interval iron metabolization. Data are mean ± standard error of the mean (SEM). P values were determined by mixed-effects model including a random effect term accounting for correlation among the measures within a same patient. (i) Time to progression of osteonecrosis between labeled and unlabeled cell transplants are not significantly different (P = 0.8, n = 16). P value was determined by log-rank test. (This figure has been reprinted with permission from Theruvath et al, Clin Ca Res 24(24): 6223-9; 2018).