MRI findings for differentiating benign and malignant soft tissue tumors: a narrative review- Part 1: diagnostic performance

This review aims to assess the diagnostic performance of MRI in distinguishing benign from malignant soft tissue tumors and evaluate conventional MRI sequences (Aim 1), as well as assess diffusion-weighted imaging (DWI) for tumor differentiation (Aim 2A) and tumor grading (Aim 2B). PubMed, Scopus, Embase, and CENTRAL were searched from inception to March 2025 using keywords like "MRI," "soft tissue sarcoma," "sensitivity," "specificity," and "diffusion." Studies with ≥ 20 cases reporting sensitivity and specificity for MRI or DWI were included for Aims 1/2A; studies on DWI for grading were included for Aim 2B. Four reviewers screened titles, abstracts, and full texts, resolving disagreements by consensus. A total of 224 studies were initially identified, and 34 were included for final analysis (Aim 1: 16, Aim 2A: 12, Aim 2B: 6). Conventional MRI (Aim 1) demonstrated sensitivities of 61-100%, specificities of 37-98%, and high negative predictive values (up to 98%) but lower positive predictive values (38-60%). DWI (Aim 2A) showed sensitivities of 73-100%, specificities of 63-94%, and lower ADC values in malignant tumors. ADC values were typically lower in high-grade tumors (Aim 2B). Overall, MRI appears to offer high diagnostic performance in excluding malignancy but lower positive predictive value and thus should be paired with histopathology or supplemental imaging. Integrating DWI into MRI protocols may enhance diagnostic accuracy by increasing specificity and confidence in distinguishing between benign and malignant tumors. ADC values from diffusion-weighted imaging were inversely correlated with tumor grade, suggesting utility in differentiating between low- and high-grade sarcomas.