Comparison of diagnostic performance between whole-body MRI with DWI and PET/MRI

Purpose: The aim of this study was to compare the diagnostic ability of whole-body (WB) MRI, with diffusion-weighted image (DWI), and integrated positron emission tomography with 2-[fluorine-18]-fluoro-2-deoxy-D-glucose (FDG-PET)-PET/MRI to detect malignant lesions and elucidate the utility and limitations of these imaging modalities for cancer screening. Methods From September 2015 to March 2016, a total of 144 patients (88 men, 56 women; mean age 59.9 years, range 9-88 years) who underwent PET/MRI including WB DWI were evaluated. All patients had histologically proven malignancies or suspected malignancies. Two radiologists interpreted the PET/MRI images in a blinded fashion. Axial T2WI, axial DWI, coronal T1-weighted images of 4-mm, 8-mm and 8 mm thicknesses, respectively, axial volumetric interpolated breath-hold examination (VIBE) of 3 mm in the lung field with 3-tesla MRI of FDG-PET/MRI were interpreted in combination with T2WI alone, T2WI + DWI, T2WI + DWI + T1WI, all MRI sequences, and all MRI and FDG-PET/MRI sequences. The primary tumors of the patients were as follows: head and neck cancer, n=29; prostate cancer, n=25; soft tissue sarcoma, n=20; lymphoma, n=15; rectal or anal cancer, n=7; melanoma, n=6; uterus cancer, n=5; osteosarcoma, n=5; lung cancer, n = 4, multiple myeloma, n=4; brain tumor, n=3; Ewing sarcoma, n=2; bladder cancer, n=2; ureter cancer, neuroblastoma, epithelioid angiosarcoma, leukemia, neuroendocrine tumor, or esophageal cancer, n=1. Patients with the following benign lesions were also included: benign bone lesion n=2; multiple neurinoma, n=2; and thymoma, hematoma, inflammatory pseudo-tumor, IgG4 related disease, vascular malformation, granulomatosis with polyangiitis, or duplication of ileum, n=1.Lesion-based sensitivity and patient-based sensitivity were calculated in each combination of the MRI sequences and FDG-PET. Histopathology, all imaging findings and follow-up scans served as a standard of reference. Results A total of 370 lesions were observed in 90 of the 144 (62.5%) patients who took part in the current study. The malignant lesions were recognized in the lymph nodes (n=139), bones (n=82), lungs (n=31), soft tissue (n=19), prostate (n=19), peritoneum (n=18), liver (n=15), neck (n=15), brain (n=8), intestines (n=6), skin (n=5), adrenal glands (n=4), uterus (n=3), eye balls (n=2), mediastinum (n=2), spleen (n=1), and pericardial sac (n=1). The lesion-based sensitivities were 63/58, 74/70, 74/70, 76/72, 93/91% and t