Graphical Abstract
Abstract
SUMMARY: Ferumoxytol, an FDA-approved treatment for iron deficiency anemia, is frequently used off-label as an MRI contrast agent. However, its unintended effects on brain MRI after anemia treatment, particularly in SWI and GRE sequences, remain insufficiently addressed. This study evaluates the optimal time interval between therapeutic ferumoxytol administration and brain MRI to minimize such effects. Analyzing 40 patients who underwent MRI within 3 months of ferumoxytol treatment, we found that 68% exhibited enhancement or signal drop on T1-weighted images and 63% showed susceptibility artifacts on SWI/GRE sequences. A 5–6 day interval between ferumoxytol administration and MRI examination is recommended to reduce these effects.
ABBREVIATIONS:
- ACR
- American College of Radiology
- AUC
- area under the curve
- ROC
- receiver operating characteristic
MRI has become an invaluable diagnostic tool in neuroradiology. While contrast-enhanced MRI studies, particularly gadolinium based, are preferred over noncontrast studies for many indications, they are often contraindicated in patients with impaired kidney function due to concerns of nephrogenic systemic fibrosis.1 The American College of Radiology (ACR) recently added ferumoxytol as an alternative contrast medium for MRI in the newest edition of its contrast manual.
Ferumoxytol, an FDA-approved therapy for iron deficiency anemia, is an iron oxide nanoparticle,2,3 which has been used off-label as an MRI contrast agent due to its T1 and T2 shortening effects.4,5 Given that patients with chronic kidney disease, especially those on dialysis, often experience iron deficiency anemia and may require intravenous iron replacement,6 it is important to recognize the potential MRI artifacts in these patients, even when noncontrast studies are ordered. In fact, ferumoxytol has an extended blood pool phase up to 15 hours and is cleared through the reticuloendothelial system, and the signal artifacts on MRI may persist up to several days or months after administration, depending on the organ.7⇓-9 For diagnostic purposes ferumoxytol is typically administered at 2–4 mg/kg whereas the standard therapeutic dosage for iron deficiency anemia treatment is 1020 mg or 14.6 mg/kg in a 70-kg adult. Consequently, the potential artifacts on MRI in patients receiving ferumoxytol for anemia treatment can be aggravated. Currently there is no established MRI protocol in these patients after ferumoxytol administration.
In this study, our goal is to determine the optimal window to obtain MRI to minimize the delayed enhancement in T1-weighted images and signal drop in SWI or GRE sequence images, particularly in brain MRI in patients who are receiving ferumoxytol therapy. We also investigate if renal function or dialysis status can affect duration of contrast effects on MR imaging.
MATERIALS AND METHODS
Study Subjects
The study was approved by the Texas Tech Health Sciences Center El Paso institutional review board and University Medical Center El Paso. In a retrospective data extraction from November 2022 to September 2023, 40 patients who received both ferumoxytol for anemia treatment and brain MR imaging within a 3-month interval were selected as study subjects. These patients were chosen based on the order of brain MR imaging following a 1-time injection of 1020 mg ferumoxytol (30 mg/mL).
Of the 40 patients included in this study, 26 (65%) were men, while 14 (35%) were women. The patient age ranges from 20 to 90 years, with a mean age of 53 years. Most patients (90%) identified as white, while the minority identified as other (7.5%), and the remaining (2.5%) identified as black. The primary reasons for the MRI examination in most patients were acute neurologic deficits or stroke (60%), followed by trauma (20%), and less common causes such as CNS infection, seizures, and other conditions such as headache, and metastasis. Clinical characteristics of study subjects are summarized in Table 1.
Clinical characteristics of study subjects
MRI Evaluation
All MRI studies were acquired by using a 3T Magnetom scanner (Siemens). The imaging protocol included sagittal T1-weighted (TE = 10 ms, TR = 2000 ms, 3 mm slice thickness), axial T2-weighted (TE = 101 ms, TR = 6100, 3 mm slice thickness), axial FLAIR (TE = 89 ms, TR = 9000 ms, 3 mm slice thickness), axial Diffusion-Weighted (TE 1 = 76 ms, TE 2 = 128 ms, TR = 6220 ms, b1 = 0 s/mmm2, b2 = 1000 s/mm2), and axial Susceptibility-Weighted (TE = 20 ms, TR = 27 ms, 1.5 mm slice thickness). Contrast-enhanced scans used gadolinium at a dose of 0.1 mmol/kg. Full MRI parameters are provided in the Supplementary Data.
MRI including T1 sagittal, SWI, or GRE sequences were evaluated independently by a board-certified neuroradiologist and 2 radiology residents. The effects of ferumoxytol on T1-weighted sequence and SWI or GRE was analyzed. T1 enhancement or signal drop in T1-weighted sequence in the brain cerebral vascular system was considered as image changes (Figs 1, and 2, respectively). Nonconspicous enhancement on T1-weighted images would be categorized as negative. Exaggerated susceptibility artifacts visualized in SWI/GRE sequences were included as SWI/GRE degradation (Fig 3). In instances where signal alterations were detected, a comparison was made with prior MRI examination, if available.
Sagittal T1-weighted images show unexpected enhancement in the cerebral veins.
Sagittal T1-weighted images in 2 different patients demonstrate signal drop early after ferumoxytol therapy.
Axial SWI shows susceptibility artifact in a patient who received ferumoxytol 3 days before MRI.
Statistical Analysis
Statistical analyses including receiver operating characteristic (ROC) analysis, univariate logistic regression analysis, and multivariate regression analysis were performed by using SPSS 29.0 (SPSS) and MedCalc 23.1.5 (MedCalc Software). A P value < .05 was considered statistically significant.
RESULTS
Among the 40 study patients, 26 (65%) showed T1 signal change and 25 (63%) showed image degradation in T2* GRE sequences. Most patients did not have a baseline MRI for comparison, and among those who had a prior MRI examination, the observed signal changes were not present at the time. The mean duration between ferumoxytol injection and brain MRI was 7 days. In the Mann-Whitney U test, the groups with positive T1 changes and degradation in T2* GRE sequences showed a significantly shorter time interval between ferumoxytol administration and MRI scanning compared with their counterparts (U = 43, P < .001 and U = 5, P < .01, respectively) (Table 2, Fig 4).
Time interval between ferumoxytol injection and MRI (days).
Summary of T1 and SWI/GRE effects results
A logistic regression analysis was also conducted to examine the impact of age, sex, and race on the likelihood of signal changes on T1-weighted imaging. The overall model was not statistically significant, χ2(3) = 7.2, P = .07. Similarly, the multivariate analysis on SWI/GRE signal change did not yield statistical significance, χ2(3) = 5.6, P = .23.
ROC curve analysis was used to predict an optimal time interval to acquire MRI postadministration without unintended T1 effects and SWI/GRE degradation. The optimal cutoff value was identified by using the maximum Youden index [sensitivity − (1−specificity)] on the ROC curve. Using a time difference of 6 days as a cutoff, the sensitivity for detecting T1 enhancement was 96% and the specificity was 80% (area under the curve [AUC] = 0.94; 95% CI, 0.75–0.96) (Fig 5). For SWI/GRE effects, by using a time difference of 5 days as a cutoff, the sensitivity was 100% and the specificity was 93% (AUC = 0.99; 95% CI, 0.89–0.98) (Fig 6).
ROC analysis was used to assess time interval between ferumoxytol administration and MRI in predicting unintended T1 effects.
ROC analysis was used to assess time interval between ferumoxytol administration and MRI in predicting unintended SWI/GRE effects.
An unexpected signal drop in T1-weighted MRI scans was observed in 5 of 27 cases (19%) among all cases with T1 signal changes. These cases had a significantly shorter interval between ferumoxytol administration and brain MR imaging compared with those with T1 enhancement (median value 1.0 day versus 2.0 days, P < .05, Mann-Whitney U test).
DISCUSSION
With a favorable nonrenal toxicity profile, ferumoxytol has been used widely as an alternative contrast agent for multiple applications.3,4,10,11 However, its potential artifacts with conventional MRI have been highlighted in several studies, especially regarding susceptibility artifacts.4,12 Depending on the organ and the imaging sequence used, the imaging effect of ferumoxytol can persist for days to months.4 In our study, ferumoxytol is found to affect T1- and T2-weighted sequences almost proportionally (35% versus 38%, respectively) with the degradation effects observed up to 7 days on T1-weighted and 5 days on T2-weighted images postadministration. It is estimated that a time interval of approximately 5–6 days between 1 injection of a standard anemia treatment dose of 1020 mg ferumoxytol and subsequent brain MRI examinations was necessary to minimize the degradation effects on T1-weighted and SWI/GRE imaging.
Ferumoxytol is known to enhance signals in T1-weighted images, typically peaking at 24 hours, but at very high concentrations, it has been reported to cause a drop in T1 signal.7 These signal behaviors are consistent with the observed T1 signal changes in our study, where T1 signal drop was associated with a shorter interval postinjection compared with T1 enhancement cases (1 day versus 2 days, respectively). This phenomenon may be similar to the decreased T1 signal observed with gadolinium. Pseudolayering of gadolinium in the bladder is an artifact caused by high concentrations of gadolinium in the urine, where T2-shortening effects outweigh T1 effects at very high concentrations.13⇓-15 Thus, when ferumoxytol concentration in the blood pool is high, susceptibility related signal loss can be exaggerated.
Limitations of our study include a relatively small number of study subjects and subjective evaluation of image changes. This is particularly relevant for SWI and GRE sequences, which are crucial for detecting small hemorrhages. In our study 2 patients with metastatic disease were found to have enhancing lesions on precontrast T1-weighted imaging that subsequently remained enhancing on postcontrast T1-weighted imaging without significant increase in signal intensity. However, these lesions also demonstrated susceptibility artifact on SWI, limiting evaluation for intralesional hemorrhage. A quantitative evaluation method could be used in future studies to further investigate the correlation between image degradation and the interval between ferumoxytol administration and MRI scanning, as well as the potential adverse effects of image degradation on clinical outcomes. Additionally, ferumoxytol clearance metabolism via the reticuloendothelial system should be evaluated to accurately measure optimal imaging acquisition time. Furthermore, due to the unique population profile at the United States-Mexico border city, the sample predominantly comprised white/Mexican descent participants, which may limit the generalizability of the findings to more diverse populations.
Beyond the technical considerations mentioned, it is important to raise awareness among radiologists and referring physicians about the prevalence of ferumoxytol use in patients with iron deficiency anemia, along with its advantages and disadvantages in MRI, to prevent misinterpretation and enhance diagnostic accuracy. In a nonurgent clinical setting, these studies may be delayed to allow for adequate clearance of ferumoxytol. Alternatively, in an acute setting, interpreting radiologists should understand the imaging effects and potential limitations with proceeding. Also, if contrast-enhanced MRI examinations are desired after ferumoxytol therapy, they can be acquired without administration of additional contrast. It is also worth noting that interactions between ferumoxytol and gadolinium-based contrast agent have been implicated according to the 2025 edition of the ACR Manual of Contrast Media.16
CONCLUSIONS
Ferumoxytol is a novel contrast medium that can be strategically used in patients with chronic kidney disease for both therapeutic and diagnostic purposes. It is estimated that a time interval of 5–6 days or more between ferumoxytol injection and brain MRI is recommended to avoid imaging degradation. Further evaluation with a larger sample with varied demographic characteristics, incorporation of hepatic clearance, quantitative analysis of signal changes, and assessment of the clinical impact of image degradation are warranted.
Footnotes
Disclosure forms provided by the authors are available with the full text and PDF of this article at www.ajnr.org.
Indicates open access to non-subscribers at www.ajnr.org
References
- Received November 12, 2024.
- Accepted after revision March 21, 2025.
- © 2025 by American Journal of Neuroradiology
Jump to section
Related Articles
Cited By...
- No citing articles found.