R2 and R2* MRI assessment of liver iron content in an undifferentiated diagnostic population with hyperferritinaemia, and impact on clinical decision making

Melissa L Craft; Morgan Edwards; Tarun Pankaj Jain; Philip Y Choi

doi:10.1016/j.ejrad.2020.109473

R2 and R2* MRI assessment of liver iron content in an undifferentiated diagnostic population with hyperferritinaemia, and impact on clinical decision making

Eur J Radiol. 2021 Feb:135:109473. doi: 10.1016/j.ejrad.2020.109473. Epub 2020 Dec 10.

Authors

Melissa L Craft¹, Morgan Edwards², Tarun Pankaj Jain³, Philip Y Choi⁴

Affiliations

¹ Department of Medical Imaging, The Canberra Hospital, 77 Yamba Dr, Garran, ACT, 2605, Australia. Electronic address: melissa.craft@act.gov.au.
² Department of Haematology, The Canberra Hospital, 77 Yamba Dr, Garran, ACT, 2605, Australia. Electronic address: claireandmog@gmail.com.
³ Department of Medical Imaging, The Canberra Hospital, 77 Yamba Dr, Garran, ACT, 2605, Australia; Universal Medical Imaging, 1/110 Giles Street, Kingston, ACT, 2604, Australia. Electronic address: tarun.jain@act.gov.au.
⁴ Department of Haematology, The Canberra Hospital, 77 Yamba Dr, Garran, ACT, 2605, Australia; The John Curtin School of Medical Research, Australian National University, 131 Garran Rd, Acton, ACT, 2601, Australia. Electronic address: Phil.Choi@act.gov.au.

PMID: 33373894
DOI: 10.1016/j.ejrad.2020.109473

Abstract

Purpose: To confirm the linear correlation between Ferriscan® R2 (1/T2 Relaxomatry) and R2* (1/T2* Relaxometry) derived 3D Gradient echo (GRE) mDIXON-Quant sequence (Philips) with simultaneous production of a proton density fat fraction (PDFF) in undifferentiated patients with hyperferritinaemia, and to prospectively determine the clinical utility of this tool in these patients by recording the impact on clinical decision-making.

Materials and methods: Participants referred to a hospital haematology outpatient clinic for investigation and management of elevated serum ferritin (two serum ferritin levels > 500 μg/L 4 weeks apart) were included in the study.

Exclusion criteria: contraindications to MRI; clinically relevant investigations for alternative causes of hyperferritinaemia pending; and terminal illness. Thirty-two participants were recruited: 27 men, 5 women. All MRIs performed at 1.5 T. For R2* quantification, 3D six echo GRE sequence (mDIXON-Quant) was acquired. R2 images were acquired over 20 min as dictated and reported by the licensee (Ferriscan®). Clinician interpretation and patient management based on R2* and liver iron content derived from R2 (LIC_R2) was recorded. Pearson's correlations, linear regression analyses, and ROC curves were calculated. P value <0.05 was considered significant.

Results: A high degree of correlation between mean R2* and LIC_R2 was observed in this novel patient population (slope ± SE of 43.35 ± 1.88 s^-1 permg/g; 95 % CI 39.5-47.2; P < 0.001; R² = 0.87). Clinical decision making was amended in 14/32 (44 %) patients with hyperferritinaemia following the disclosure of R2* results to clinicians, compared with serum ferritin alone. Liver biopsy was avoided in one patient based on LIC_R2 and R2*. Unrecognised hepatic steatosis was diagnosed in one patient from the PDFF map.

Conclusion: We have confirmed the linear correlation between R2 and R2* in a real-world diagnostic population with hyperferritinaemia. Non-invasive assessment of liver iron content (LIC) by R2 and R2* MRI is a useful clinical tool and alters management in these patients.

Keywords: Abdominal magnetic resonance imaging; Ferriscan; Hyperferritiaemia; Iron overload; Liver; Proton density fat fraction; Relaxometry.

MeSH terms

Clinical Decision-Making
Female
Humans
Iron Overload*
Iron*
Liver / diagnostic imaging
Magnetic Resonance Imaging
Male

Substances

Iron