Whole-body MRI in oncology: can a single anatomic T2 Dixon sequence replace the combination of T1 and STIR sequences to detect skeletal metastasis and myeloma?

• DOI: 10.1007/s00330-022-09007-8
• Publication Date: 2022-08-04T10:03:03Z
• Authors: Ophelye Chiabai; Sandy Van Nieuwen...; Marie-Christiane ...; Bertrand Tombal; Frank Peeters; Joris Wuts; Perrine Triqueneaux; Patrick Omoumi; Thomas Kirchgesner; Nicolas Michoux; Frédéric E. Lecouvet
• Keywords: Neuroradiology; Repeatability; Whole body imaging

To compare the diagnostic accuracy of a single T2 Dixon sequence to combination T1+STIR as anatomical sequences used for detecting tumoral bone marrow lesions in whole-body MRI (WB-MRI) examinations.Between January 2019 and 2020, seventy-two consecutive patients (55 men, 17 women, median age = 66 years) with solid (prostate, breast, neuroendocrine) cancers at high risk metastasis or proven multiple myeloma (MM) prospectively underwent WB-MRI examination including coronal T1, STIR, axial diffusion-weighted imaging sequences. Two radiologists independently assessed fat+water reconstructions from sequence. The reference standard was established by consensus reading concurrent available baseline 6 months. Repeatability reproducibility scores (presence semi-quantitative count lesions), image quality (SNR: signal-to-noise, CNR: contrast-to-noise, CRR: contrast-to-reference ratios), characteristics (Se: sensitivity, Sp: specificity, Acc: accuracy) were per-skeletal region per-patient.Repeatability least good regardless score, region, protocol (0.67 ≤ AC1 0.98). CRR higher on fat compared T1 (p < 0.0001) water STIR 0.0128). In per-patient analysis, Acc than that senior reader (Acc +0.027 [+0.025; +0.029], p lower junior -0.029 [-0.031; -0.027], 0.0001).A offers similar recommended can be skeletal screening oncology, allowing significant time-saving.• Replacement anatomic + drastically shortens time without loss accuracy. • A based inter-reader agreement ratio protocol. Differences between two protocols are marginal (+ 3% favor reader; -3% against reader).