Purpose: The purpose of this prospective study was to assess value of imaging diagnosis in differentiation of benign from malignant soft-tissue lesions, ability to make accurate preoperative staging and to make diagnostic algorithm.
Material and methods: One hundred consecutive patients (53 male, 47 female, median age 51 years) with musculoskeletal soft tissue mass underwent ultrasonography(US), computed tomography(CT) and magnetic resonance(MRI). Benign lesions were 70(53 benign tumors and 17 tumor-like lesions) and 30 were malignant lesions (26 malignant tumors and 4 metastases). Fifteen imaging parameters on US, CT and MR were evaluated. The diagnosis was based on histopathologic findings from core needle biopsy and surgical specimens. Multivariate logistic regression analysis was used to identify the best combination of imaging parameters in all used methods, predicting malignancy. Results: For predicting primary malignant soft tissue musculoskeletal tumor, US had sensitivity 80,8%, specificity 77% and accuracy 78%. For benign tumors the accuracy was 94%. The correlation between imaging preoperative staging of malignant tumors was 63,6%. Following US parameters for predicting malignancy showed statistically significant difference(p<0,01): partial capsule or uncapsulated lesions, inhomogeneous structure, presence of necrosis and contact with vascular structures.
Conclusion: US had satisfactory contribution, but MRI had highest accuracy in imaging diagnosis of musculoskeletal lesions. A combination of individual parameters improved the differentiation between benign and malignant soft-tissue musculoskeletal lesions. Imaging diagnostic algorithm gives direction to evaluate musculoskeletal lesions in order not to make unnecessary examinations as well as not to miss malignant ones.