Whole-body MRI of paediatric malignant tumours: comparison with conventional oncological imaging methods

Purpose: To evaluate the usefulness of whole-body (WB) MRI for detecting metastases from paediatric malignant tumours in comparison with conventional oncological imaging methods. Materials and methods: Using a 1.5-T system, a coronal short tau inversion recovery (STIR) sequence was obtained in all patients. In addition, sagittal fat-suppressed T2-weighted, sagittal STIR, or coronal fat-suppressed pre-contrast and post-contrast T1-weighted sequences were performed. Patients who underwent WB MRI and conventional oncological imaging within 15 days were enrolled in the study. In total, 58 bone scintigraphies, 26 iodine-123 (¹²³I) meta-iodo-benzylguanidine (MIBG) scintigraphies, and 48 CT scans were available for comparison in 36 patients (median age 3.5 years; 21 boys, 15 girls) who underwent 82 WB MRI examinations. Skeletal and extraskeletal metastases were evaluated for a variety of tumour types. Results: Concordance rate of WB MRI between two readers was 74%. In detecting metastases, WB MRI had higher sensitivity (99%) and PPV (94%) than bone scintigraphy (26 and 76%, respectively). In detecting skeletal metastases, WB MRI revealed higher sensitivity (100%) than ¹²³I-MIBG scintigraphy (25%) and CT (10%). In contrast, WB MRI showed lower PPV in detecting skeletal and extraskeletal metastases (8 and 57%, respectively) than ¹²³I-MIBG scintigraphy (100%), and lower sensitivity (60%) in detecting extraskeletal metastases than CT (100%). In 2 of 11 untreated patients, tumour staging was upgraded from stage 3 to 4 according to WB MRI findings. In 3 patients, WB MRI revealed early treatment responses (<1 year) of skeletal metastases. Conclusions: WB MRI can substitute for bone scintigraphy in detecting skeletal metastases of paediatric malignant tumours, and it is useful in evaluating initial tumour staging and early treatment responses. However, it still has only a complementary role in detecting extraskeletal metastases.