Ionising radiation-free whole-body MRI versus ¹⁸F-fluorodeoxyglucose PET/CT scans for children and young adults with cancer: a prospective, non-randomised, single-centre study

Summary

Background

Imaging tests are essential for staging of children with cancer. However, CT and radiotracer-based imaging procedures are associated with substantial exposure to ionising radiation and risk of secondary cancer development later in life. Our aim was to create a highly effective, clinically feasible, ionising radiation-free staging method based on whole-body diffusion-weighted MRI and the iron supplement ferumoxytol, used off-label as a contrast agent.

Methods

We compared whole-body diffusion-weighted MRI with standard clinical ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET/CT scans in children and young adults with malignant lymphomas and sarcomas. Whole-body diffusion-weighted magnetic resonance images were generated by coregistration of colour-encoded ferumoxytol-enhanced whole-body diffusion-weighted MRI scans for tumour detection with ferumoxytol-enhanced T1-weighted MRI scans for anatomical orientation, similar to the concept of integrated ¹⁸F-FDG PET/CT scans. Tumour staging results were compared using Cohen's κ statistics. Histopathology and follow-up imaging served as the standard of reference. Data was assessed in the per-protocol population. This study is registered with ClinicalTrials.gov, number NCT01542879.

Findings

22 of 23 recruited patients were analysed because one patient discontinued before completion of the whole-body scan. Mean exposure to ionising radiation was 12·5 mSv (SD 4·1) for ¹⁸F-FDG PET/CT compared with zero for whole-body diffusion-weighted MRI. ¹⁸F-FDG PET/CT detected 163 of 174 malignant lesions at 1325 anatomical regions and whole-body diffusion-weighted MRI detected 158. Comparing ¹⁸F-FDG PET/CT to whole-body diffusion-weighted MRI, sensitivities were 93·7% (95% CI 89·0–96·8) versus 90·8% (85·5–94·7); specificities 97·7% (95% CI 96·7–98·5) versus 99·5% (98·9–99·8); and diagnostic accuracies 97·2% (93·6–99·4) versus 98·3% (97·4–99·2). Tumour staging results showed very good agreement between both imaging modalities with a κ of 0·93 (0·81–1·00). No adverse events after administration of ferumoxytol were recorded.

Interpretation

Ferumoxytol-enhanced whole-body diffusion-weighted MRI could be an alternative to ¹⁸F-FDG PET/CT for staging of children and young adults with cancer that is free of ionising radiation. This new imaging test might help to prevent long-term side-effects from radiographic staging procedures.

Funding

Thrasher Research Fund and Clinical Health Research Institute at Stanford University.

Introduction

Accurate staging of malignant tumours in children and adolescents is crucial, because the presence and location of tumours throughout the body determines therapy management and prognosis.¹ Thus, the use of CT and ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET/CT staging examinations in paediatric oncology patients has grown rapidly during the past several years.² However, CT and ¹⁸F-FDG PET/CT examinations are associated with substantial exposure to ionising radiation.² Even with child-adapted low-dose protocols, patients undergoing a single ¹⁸F-FDG PET/CT examination are exposed to ionising radiation in the order of 10–20 mSv, which is equivalent to roughly 700–750 chest radiographs.² By comparison, the yearly background dose from natural radiation exposure amounts to roughly 3 mSv.³ According to David Brenner and colleagues,³ direct epidemiological evidence from human population studies shows that doses of ionising radiation greater than 50–100 mSv (protracted exposure) or 10–50 mSv (acute exposure) increases the risk of some cancers. This risk is particularly worrying with regard to young patients because they are more susceptible to the effects of ionising radiation than older patients, and they live long enough to encounter secondary cancers.1, 3

As per the BEIR report VII,⁴ exposure to ionising radiation in early childhood causes roughly a tripling in lifetime cancer risk by comparison with a person exposed above the age of 30 years. Pearce and colleagues⁵ reported that cumulative doses from diagnostic CT scans, applied for tumour staging in children, almost tripled the risk of secondary leukaemia and brain cancer later in life. Mathews and colleagues⁶ identified a 24% increase in cancer incidence in 11 million Australians who underwent CT scans for cancer staging during childhood and adolescence. In paediatric patients and young adults with Hodgkin's disease, over 90% of patients survive 10 years, whereas only 58% of patients older than 60 years survive for 5 years.⁷ Therefore, radiation-induced mortality and risk of development of secondary cancers is more worrying in young patients. If the techniques for disease diagnosis bear a risk of secondary tumour formation later in life, this counteracts the goal of achieving long-term survival. This conundrum might be resolved by avoiding unnecessary exposure to ionising radiation associated with diagnostic imaging approaches.

MRI provides a promising alternative for whole-body staging that is free of ionising radiation.⁸ However, initial approaches to whole-body MRI were hampered by limited tumour-to-background contrast and long scan times of 1·5–2·0 h.⁹ The introduction of rapid, diffusion-weighted sequences improved tumour delineation and image acquisition times,⁹ with promising initial results in adults¹⁰ and children.⁹ However, despite these recent advances, three major limitations have prevented the adoption of whole-body diffusion-weighted MRI for tumour staging in clinical practice. First, the spleen and bone marrow show similar signal compared with tumours on diffusion-weighted sequences.¹¹ This can confound tumour detection in those organs, creating the potential for undertreatment and therapy relapse.¹² Second, T1-weighted MRI scans for anatomical orientation need contrast-agent enhancement. However, the enhancement gained with standard magnetic resonance contrast agents does not persist for the duration of a whole-body scan. Third, diffusion-weighted MRI sequences provide limited background information, similar to ¹⁸F-FDG PET scans without the CT component. To our knowledge, MRI techniques for tumour detection (whole-body diffusion) have not been integrated with MRI techniques for anatomical orientation, in accordance with the concept of integrated ¹⁸F-FDG PET/CT scans.

Our aim was to address these challenges and create a highly effective, clinically feasible, ionising radiation-free staging method. We established novel whole-body diffusion-weighted MRI scans by combining the use of clinically applicable iron oxide nanoparticles for improved delineation of tumours in spleen and bone marrow on diffusion-weighted MRI scans; the use of iron oxide nanoparticles for long lasting positive vessel enhancement on T1-weighted, anatomical magnetic resonance scans; and fusion of the two approaches for simultaneous display of highly sensitive tumour maps on a high-resolution anatomical background. We postulated that the resultant novel whole-body diffusion-weighted MRI scans would provide equal diagnostic accuracy compared with standard ¹⁸F-FDG PET/CT scans.