Correlation of magnetic resonance imaging morphological features and diffusion-weighted imaging to soft tissue sarcoma grading

Abstract Background Soft tissue sarcomas (STSs) are a diverse group of rare malignant diseases, with over 50 histologic subtypes and a significant mortality rate. Their treatment decisions are strongly influenced by the histopathologic grade of the tumor. Magnetic resonance imaging (MRI) is the imaging modality of choice for characterizing soft tissue tumors, precise tumor localization, assessment of tumor relationship to major anatomic landmarks, and for the evaluation of tumor vascularization. However, the accurate diagnosis is invariably based on histopathologic analysis. This study aimed to correlate different MRI morphological features and DWI criteria to the histopathological grading of STSs. Results The study included 46 STSs, 71.7% high-grade sarcoma (26.1% grade II and 45.6% grade III), and 28.3% low-grade (I) sarcoma according to the French federation of cancer centers sarcoma system. The most common histopathological subtypes were synovial sarcoma, myxoid liposarcoma, and undifferentiated pleomorphic sarcoma. Ill-defined margins, ≥ 75% lesions’ heterogeneity, tumoral necrosis, and peritumoral enhancement showed significant correlation to high-grade STS. Peritumoral enhancement was the most logistic feature to predict high-grade lesions. High-grade STSs had significantly lower apparent diffusion coefficient values with a cut-off value of 0.75 (× 10 −3 mm 2 /s) with 81.3% sensitivity and 92.9% specificity in distinguishing low from high-grade lesions. Conclusion Among the MRI criteria, peritumoral enhancement is the most reliable feature to predict high-grade STS.