MRI Features as Prognostic Factors in Myxofibrosarcoma: Proposal of MRI Grading System

Abstract

Rationale and objectives: Myxofibrosarcoma (MFS) is a common soft tissue sarcoma in the elderly patients with both clinical and magnetic resonance imaging (MRI) peculiar features: very high recurrence rate, relatively low risk of distant metastases. On MRI it shows an infiltrative pattern ("tail sign") and high myxoid matrix content with water-like appearance on fluid-sensitive sequences. Due to these unusual characteristics, we propose a specific MRI grading system to stratify the risk of local recurrence (LR) and offer other prognostic information.

Materials and methods: Two expert radiologists retrospectively and blindly reviewed preoperative MRI of 150 patients affected by MFS of the extremities treated at a single Institution. Myxoid matrix component and contrast enhancement of the tumor were evaluated and graded with a semiquantitative method. The presence of an infiltrative pattern, the depth of the tumor (deep and/or superficial) and tumor sizes were also recorded. MRI features were analyzed separately and correlated to LR risk, sarcoma specific survival and distant metastases rate. Then, according to the statistical significance of the correlation between MRI features and prognosis a 3-grade scoring system was proposed and evaluated to assess the risk of LR.

Results: Mean age was 66.1 ± 14.4 years; mean follow-up was 16 ± 28.3 months. The MRI features most associated with higher risk of LR resulted to be: lesion sizes (both volume and maximum diameter with a cut-off of 20 cm - p = 0.01), the "tail sign" (p = 0.045), and high myxoid matrix content with MRI water-like appearance (p = 0.0493). Ninety-four patients (94 of 150- 62.7%) were grade 1, 33 (22.0%) grade 2, and 23 (15.3%) grade 3. Interobserver agreement was substantial with K= 0.779 (95%CI 0.685-0.874). Higher grades of MRI grading system proposed were significantly associated with an increased LR risk, hazard ratio = 2.031 (95%CI 1.366-3.019; p < 0.001).

Conclusion: This is the largest series evaluating MRI features as prognostic factors for MFS. The MRI grading system proposed is significantly able to stratify the risk of LR in MFS of the extremities. The system is applicable to all the standard MRI studies protocols, might help in surgical planning, and may offer prognostic information.

Keywords: Local Neoplasm Recurrence; Magnetic Resonance Imaging; Prognosis; Sarcoma.