OBJECTIVE: The purpose of this study is to assess the role of whole-body (WB) DWI as a screening modality for the detection of metastases and to compare it to conventional cross-sectional imaging modalities or nuclear scintigraphy in a population with various histopathologic malignancies. SUBJECTS AND METHODS: WB DWI and conventional imaging (CT, MRI, or scintigraphy) were performed for patients with known malignancies for metastatic workup, and these patients were followed up for a period of 1 year. Two radiologists assessed WB DW images separately, and conventional images were assessed by the senior radiologist. The metastatic lesions were classified into four regions: liver, lung, skeletal system, and lymph nodes. The reference standard was considered on the basis of histopathologic confirmation or clinical follow-up of the metastatic lesions. RESULTS: WB DWI was slightly inferior to conventional imaging modalities for the detection of hepatic metastases (sensitivity, 86.6% vs 93.3%; specificity, 91.6% vs 95.8%; and accuracy, 89.7% vs 94.8%) and skeletal metastases (sensitivity, 81.8% vs 89.4%; specificity, 86.4% vs 94.3%; and accuracy, 85.2% vs 93.0%); however, the differences were not statistically significant (p = 0.625 for hepatic metastases and p = 0.0953 for skeletal metastases, McNemar test). WB DWI was statistically significantly inferior to conventional imaging for the detection of lymph node metastases (sensitivity, 74.0% vs 81.5%; specificity, 87.9% vs 90.1%; accuracy, 81.4% vs 86.0%; p = 0.0389). WB DWI was statistically significantly inferior to conventional imaging for the detection of pulmonary metastases (sensitivity, 33.3% vs 100.0%; specificity, 90.9% vs 100.0%; accuracy, 60.8% vs 100.0%; p = 0.045). CONCLUSION: WB DWI can be used for screening hepatic and skeletal metastases, but its reliability as the sole imaging sequence for the detection of lymph nodal and pulmonary metastases is poor and, at present, it cannot replace conventional imaging modalities.
OBJECTIVE: The purpose of this study is to assess the role of whole-body (WB) DWI as a screening modality for the detection of metastases and to compare it to conventional cross-sectional imaging modalities or nuclear scintigraphy in a population with various histopathologic malignancies. SUBJECTS AND METHODS: WB DWI and conventional imaging (CT, MRI, or scintigraphy) were performed for patients with known malignancies for metastatic workup, and these patients were followed up for a period of 1 year. Two radiologists assessed WB DW images separately, and conventional images were assessed by the senior radiologist. The metastatic lesions were classified into four regions: liver, lung, skeletal system, and lymph nodes. The reference standard was considered on the basis of histopathologic confirmation or clinical follow-up of the metastatic lesions. RESULTS: WB DWI was slightly inferior to conventional imaging modalities for the detection of hepatic metastases (sensitivity, 86.6% vs 93.3%; specificity, 91.6% vs 95.8%; and accuracy, 89.7% vs 94.8%) and skeletal metastases (sensitivity, 81.8% vs 89.4%; specificity, 86.4% vs 94.3%; and accuracy, 85.2% vs 93.0%); however, the differences were not statistically significant (p = 0.625 for hepatic metastases and p = 0.0953 for skeletal metastases, McNemar test). WB DWI was statistically significantly inferior to conventional imaging for the detection of lymph node metastases (sensitivity, 74.0% vs 81.5%; specificity, 87.9% vs 90.1%; accuracy, 81.4% vs 86.0%; p = 0.0389). WB DWI was statistically significantly inferior to conventional imaging for the detection of pulmonary metastases (sensitivity, 33.3% vs 100.0%; specificity, 90.9% vs 100.0%; accuracy, 60.8% vs 100.0%; p = 0.045). CONCLUSION: WB DWI can be used for screening hepatic and skeletal metastases, but its reliability as the sole imaging sequence for the detection of lymph nodal and pulmonary metastases is poor and, at present, it cannot replace conventional imaging modalities.
Authors: Stuart A Taylor; Susan Mallett; Anne Miles; Stephen Morris; Laura Quinn; Caroline S Clarke; Sandy Beare; John Bridgewater; Vicky Goh; Sam Janes; Dow-Mu Koh; Alison Morton; Neal Navani; Alfred Oliver; Anwar Padhani; Shonit Punwani; Andrea Rockall; Steve Halligan Journal: Health Technol Assess Date: 2019-12 Impact factor: 4.014