Whole body diffusion-weighted MRI in detection of metastasis and lymphoma: a prospective longitudinal clinical study

2020 
Whole-body diffusion-weighted magnetic resonance imaging (WB-DWI-MRI) is an emerging tool that has an increasing role in the diagnosis of metastasis and lymphoma. This is a longitudinal study in actual clinical settings designed to assess WB-DWI-MRI in detection of tumor spread. The study included all patients who were referred to Radiology Department, during the period from June 2016 till May 2018, with either a known primary tumor (either laboratory, radiologically, or histologically proven, of any type, affecting any organ) or with biopsy-proven lymphoma of any subtype, affecting any organ. All patients underwent WB coronal T1-weighted, STIR, axial T2-weighted, and DWI-MRI examinations before commencing any treatment with curative intent. The body was divided into lymph nodes (LNs), skeletal system, and organs (brain, lung, and liver). Patients were followed up till the nature of the lesion(s) was confirmed (clinically, radiologically, or histologically). The study included 46 patients; 27 patients had metastases and 19 had lymphomas. Sensitivities, specificities, and accuracies for LN detection were 77%, 85%, and 83%; for skeletal metastasis were 88%, 94%, and 92%; for brain lesions were 78%, 95%, and 91%; and for lung lesion were 64%, 88%, and 76%, respectively. As for the liver, all lesions were correctly identified and did not miss any lesion with accuracy of 100%. Overall, 1739 lesions were discovered in 1271 regions out of 3818 examined regions with overall sensitivity, specificity, and accuracy of 86%, 92%, and 90% respectively. The diagnostic performance of WB-DWI-MRI is variable among different anatomical sites. It has good performance in diagnosis of some organs as liver, bone marrow, and some LNs regions as porta-hepatis. It has a less diagnostic performance in the lung, and LNs located in cervical, mediastinum, supraclavicular, and mesenteric regions.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    30
    References
    1
    Citations
    NaN
    KQI
    []