Diffusion-weighted imaging for the cellularity assessment and matrix characterization of soft tissue tumour.

PURPOSE: To evaluate whether apparent diffusion coefficient (ADC) of diffusion-weighted imaging (DWI) is able to investigate the histological features of soft tissue tumours. METHODS: We reviewed MRIs of soft tissue tumours performed from 2012 to 2015 to calculate the average ADCs. We included 46 patients (27 male; mean age: 57 years, range 12-85 years) with histologically proven soft tissue tumours (10 benign, 2 intermediate 34 malignant) grouped into eight tumour type classes. An experienced pathologist assigned a semi-quantitative cellularity score (very high, high, medium and low) and tumour grading. The t test, ANOVA and linear regression were used to correlate ADC with clinicopathological data. Approximate receiver operating characteristic curves were created to predict possible uses of ADC to differentiate benign from malignant tumours. RESULTS: There was a significant difference (p < 0.01) in ADCs between these three groups excluding myxoid sarcomas. A significant difference was also evident between the tumour type classes (p < 0.001), grade II and III myxoid lesions (p < 0.05), tumour grading classes (p < 0.001) and cellularity scores classes (p < 0.001), with the lowest ADCs in the very high cellularity. While the linear regression analysis showed a significant relationship between ADC and tumour cellularity (r = 0.590, p ≤ 0.05) and grading (r = 0.437, p ≤ 0.05), no significant relationship was found with age, gender, tumour size and histological subtype. An optimal cut-off ADC value of 1.45 × 10-3 mm2/s with 76.8% accuracy was found to differentiate benign from malignant tumours. CONCLUSIONS: DWI may offer adjunctive information about soft tissue tumours, but its clinical role is still to be defined.