The Value Of DWI-ADC In Differential Diagnosis Between Lipoma And Liposarcoma

Objective: To compare the image characteristic of Diffusion Weighted Imaging(DWI) and Apparent Diffusion Coefficient(ADC) map of lipoma and liposarcoma. To compare the ADC values of lipoma and well-differentiated liposarcoma. To explore the potential value of DWI-ADC in distinguishing liposarcoma from lipoma.Materials and methods:1.Patient: We chosed 33 cases(Aged 19-80 years, with an median age of 56 years) of fatty soft-tissue masses in limbs and trunk for surgical treatment from June 2008 to February 2015. O f the 33 fatty soft-tissue tumor cases, there were 12 cases of lipoma and 21 cases of liposarcoma. O f the 21 liposarcoma cases, there were 10 cases of well-differentiated liposarcoma, 5 cases of myxoid liposarcoma, 4 cases of dedifferentiated liposarcoma, and 2 cases of pleomorphic liposarcoma. All cases were confirmed by surgical pathology.2.MR technology and image post-processing: 33 cases were scanned routine MRI and DWI using GE Signa HDxt 1.5T MR. DWI(Single Shot SE-EPI;b=0,600s/mm2) were scanned with frequency selective saturation technology to suppress fat. DWI were sent to GE-ADW 4.4 workstation, Functool software was used to post-process the original image and generate ADC map. Volume Rendering software was used to obtain such quantitative parameters as ADC mean and ADC(no- fat).3. Observational indexes:(1)DWI and ADC map manifestation: Presence or absence of high signal on DWI and low signal on ADC map;(2) ADC values: ADC mean、ADC(no- fat).4. Statistical methods:(1)Fisher’s exact test was used to compare the difference of DWI and ADC map manifestation between lipoma and liposarcoma.(2)Wilcoxon rank test was used to compare the difference of ADC values between lipoma and well-differentiated liposarcoma;Receiver operating characteristic(ROC) analysis was performed to assess an optimal threshold value for distinguishing well-differentiated liposarcoma from lipoma. The area under the ROC curve between ADC mean and ADC(no-fat)were compared using z test.Results:1. DWI and ADC map manifestation comparison of lipoma and liposarcoma: 100%(12/12)of the lipoma were absence of high signal on DWI and low signal on ADC map, 62%(13/21)of liposarcoma were presence of high signal on DWI and low signal on ADC map, 38%(8/21)cases of liposarcoma were absence of high signal on DWI and low signal on ADC map.There were significant differences between lipoma and liposarcoma in the presence or absence of high signal on DWI and low signal on ADC map(P=0.001).2. ADC values comparison of lipoma and well-differentiated liposarcoma: ADC mean and ADC(no- fat)of lipoma were lower than that of well-differentiated liposarcoma. There were significant differences between lipoma and well-differentiated liposarcoma in ADC mean and ADC( no-fat)(P < 0.05). A cut-off value of ≥1962.0?10-6mm2/s for ADC(no-fat) in differentiating liposarcoma from lipoma with a specificity of 100%, a sensitivity of 80%. There was no significant difference in the AUC between ADC mean and ADC(no- fat) for distinguishing well-differentiated liposarcoma from lipoma(P? 0.05).Conclusion:1. DWI and ADC map manifestation can be used in the differential diagnosis between lipoma and liposarcoma. 100%(12/12)of the lipoma were absence of high signal on DWI and low signal on ADC map, 62%(13/21)of liposarcoma were presence of high signal on DWI and low signal on ADC map.2. The ADC values can be used to identify the lipoma and well-differentiated liposarcoma, the ADC mean and ADC(no-fat) of lipoma were significantly lower than that of well-differentiated liposarcoma.