Added value of diffusion-weighted imaging to conventional MRI for predicting fascial involvement of soft tissue sarcomas.

OBJECTIVES: This study was conducted to evaluate the added value of diffusion-weighted imaging (DWI) to conventional magnetic resonance imaging (MRI) for predicting fascial involvement of soft tissue sarcomas located in close proximity to fascial boundaries.

METHODS: This retrospective study included 29 patients with surgically resected soft tissue sarcomas located in proximity to deep fascia and with a curvilinear tail-like hyperintensity in the adjacent fascia on T2-weighted images. All patients underwent conventional MRI and DWI at 3.0 T and had detailed histologic reports on involvement of fascia. Two musculoskeletal radiologists with 21 and 1 year of experience independently reviewed conventional MRI and conventional imaging with added DWI. Readers scored their confidence for tumor involvement of fascia using a three-point scale. Diagnostic performance (area under the curve [Az]) of the two MRI sets was assessed with receiver-operating characteristic curve analysis.

RESULT: Fascial involvement was present in 22/29 patients (75.9%). Both readers showed improvement in diagnostic performance with the addition of DWI (Az, from 0.545 to 0.792 and from 0.646 to 0.792 for reader 1 and reader 2, respectively). Adding DWI did not improve sensitivity or specificity for either reader (p > 0.05). Interobserver agreement for the confidence scores improved from fair to moderate with the addition of DWI (κ, from 0.390 to 0.560).

CONCLUSIONS: Adding DWI to conventional MRI improved diagnostic performance on prediction of fascial involvement of soft tissue sarcomas located in proximity to fascia, without significant improvement in sensitivity or specificity.

KEY POINTS: • Adding DWI to conventional MRI improved readers' confidence level for the prediction of fascial involvement of soft tissue sarcomas that are close to the deep fascia. • Addition of DWI also improved interobserver agreement. • Conversely, compared with conventional MRI, adding DWI did not significantly improve the sensitivity or specificity for the detection of fascial involvement.