The feasibility of apparent diffusion coefficient (ADC) histogram analysis in differentiating recurrences from...

To explore the feasibility of quantitative apparent diffusion coefficient (ADC) histogram values in the differentiation of recurrences and postsurgical changes of soft tissue sarcoma (STS).

The clinicopathological and magnetic resonance (MR) imaging (with diffusion-weighted imaging) data of 64 patients (37 with recurrences of STS and 27 with postsurgical changes) were prospectively reviewed. The MR features and ADC histogram values were compared between recurrences and postsurgical changes of STS.

Recurrences of STS were larger and commonly accompanied by internal necrosis and neurovascular encasement. There were statistically significant differences between the mean ADC value, the 10^th ADC value, 25^th ADC value, 50^th ADC value, 75^th ADC value, 90^th ADC value, and skewness value between recurrences and postsurgical changes. The cut-off values were ≤1.204 × 10^-3mm²/s, ≤0.835 × 10^-3mm²/s, ≤1.095 × 10^-3mm²/s, ≤1.238 × 10^-3mm²/s, ≤1.352 × 10^-3mm²/s, ≤1.637 × 10^-3mm²/s, and >-0.43, respectively. The 10^th ADC value was the independent predictive factor with an odds ratio (OR) value of 0.99. The sensitivity, specificity, accuracy, and the area under the curve (AUC) of logistic regression were 78.38%, 100%, 84.37%, and 0.937 (95% confidence interval [CI]: 0.847-0.982), respectively.

The ADC histogram values helps in differentiating recurrences from postsurgical changes of STS, and the 10^th ADC value can serve as an independent predictive factor.