Whole-body magnetic resonance imaging (WB-MRI) for cancer screening: recommendations for use

Abstract

Whole-body magnetic resonance imaging (WB-MRI) is currently recommended for cancer screening in adult and paediatric subjects with cancer predisposition syndromes, representing a substantial aid for prolonging health and survival of these subjects with a high oncological risk. Additionally, the number of studies exploring the use of WB-MRI for cancer screening in asymptomatic subjects from the general population is growing. The primary aim of this review was to analyse the acquisition protocols found in the literature, in order to identify common sequences across published studies and to discuss the need of additional ones for specific populations. The secondary aim of this review was to provide a synthesis of current recommendations regarding the use of WB-MRI for cancer screening.

Keywords: Cancer screening; Cancer-related syndromes; Diffusion-weighted imaging; Magnetic resonance imaging; Oncology; Whole-body MRI.

Fig. 1

Core protocol. This image summarizes the main pulse sequences included in the WB-MRI protocol for cancer screening. The study should cover from head to pelvis. T1 weighted images are acquired with a gradient recalled echo (GRE) sequence, preferably with Dixon technique, to obtain four different sets of images (in-phase, opposed-phase, water-only, fat-only) and to allow the calculation of the relative fat-fraction (rF%) map. T2 images are acquired with a single-shot fast spin echo (FSE) sequence. Diffusion-weighted images (DWI) should be acquired with at least two b values in order to obtain the corresponding apparent diffusion coefficient (ADC) map. Maximum intensity projections (MIP) should be reconstructed from the highest b value DWI images

Fig. 2

Extensions to the core protocol. This picture summarizes possible complements to the core protocol. When assessment of the spine is required, sagittal short tau inversion recovery (STIR) is used. In subjects with a high risk of vertebral tumours, a sagittal T1 weighted turbo spin echo (TSE) sequence can be additionally performed. When the WB-MRI protocol includes the lower limbs, such as in subjects with Li Fraumeni syndrome, all sequences in the core protocol are extended to the feet. When there is an increased risk of central nervous system (CNS) tumours, a dedicated brain sub-protocol is performed, with multiple sequences and with contrast administration (+ c). In subjects with a low risk of CNS tumours, the assessment of the brain can be improved with a short brain protocol, including fluid attenuated inversion recovery (FLAIR) sequences. Single breath-hold T1 weighted gradient recalled echo (GRE) sequences can be performed for the assessment of the lungs. In subjects with neurofibromatosis, the STIR sequence should be performed in either the axial or the coronal plane, covering from neck to feet, for facilitating the detection of peripheral nerve sheath tumours (arrows)

Fig. 3

Case example. Images of a 29-year-old woman with Li-Fraumeni syndrome with prior history of giant cell fibroblastoma of the groin (13 years before), Paget disease of the right breast and grade-3 ductal intraepithelial neoplasia (DIN-3) of the right breast (8 years before). The patient underwent a first screening with WB-MRI, which revealed multiple abnormal findings. The high b-value maximum intensity projection displayed in lateral view with inverted grayscale (A) revealed three hyper-intense lesions in the pelvis. Firstly, a 8-cm mass highly suspicious for cancer was detected in the right gluteus (arrow in A and B), showing hyper intense appearance in high b-value images (top row, A and B), cystic areas in T2 weighted images (middle row, A and B) and irregular contrast enhancement in delayed post-contrast T1 weighted (W) Dixon images (bottom row, A and B). Secondly, a solid lesion with irregular shape was seen adjacent to the right femoral vessels (arrowhead in A and C), with hyper-intense appearance in high b-value images, heterogeneous signal in T2 W images and strong enhancement in post-contrast T1 W Dixon images (A and C, top, middle and bottom row, respectively). The finding was reported as strongly suspicious for local recurrence of fibroblastoma. Thirdly, an enlarged femoral lymph node was seen in the right thigh (dashed arrowhead in A). The patient underwent surgical resection of the suspicious lesion in the gluteus and dissection of the right groin, with histopathological diagnosis of high-grade sarcoma of the gluteus and local recurrence of fibroblastoma. Metastasis from high-grade sarcoma was diagnosed in the enlarged femoral lymph node. Two other focal lesions were detected, which were not visible in high b-value images (D and E, top row). In D, a solid, rounded lesion was seen the VII segment of the liver, with high signal intensity in T2 W images (D and E, middle row) and evidence of intralesional fat in the relative fat-fraction (rF%) map (D, bottom row). The lesion showed similar contrast enhancement compared to surrounding parenchyma in T1 W Dixon images (not shown). Although not suspicious for malignant cancer, the lesion was not visible in a prior MRI study; therefore, a percutaneous liver biopsy was performed, with benign findings suggestive of chronic inflammation, macrovescicular steatosis and ductal hyperplasia. Follow-up at seven months distance showed stable findings. In E, a lobulated bone lesion was seen in the right ilium, with minimal remodelling of the cortical bone and hyper-intense content in T2 W images (middle row), and intralesional fat content below 5% in the rF% map. The lesion was not visible in a prior MR study, and in the suspect of malignancy, a percutaneous biopsy was performed, which revealed sparse foci of epithelial tumoral cells, with immunohistochemical findings compatible with metastasis from occult breast cancer. Subsequent mammography and ultrasonography revealed no abnormal breast lesions and the patient is at present under strict follow-up