Detecting residual soft tissue sarcoma after unplanned excision; model-free analysis of dynamic contrast-enhanced MRI at short-term follow-up

Abstract

Objectives To evaluate diagnostic utility of additional DCE-MRI for detecting residual soft tissue sarcomas (STS) after unplanned excision (UPE). Methods We retrospectively evaluated 32 patients with UPE of STS, followed by conventional MRI with DCE-MRI and wide excision (WE), between November 2019 and January 2022. Residual tumors on conventional MRI were categorized into three groups: Lesion-type-0, no abnormal enhancement, Lesion-type-1, an indeterminate lesion, and Lesion-type-2, a definite enhancing nodule. On DCE-MRI, ROIs were manually placed on enhancing areas of suspected residual tumor. The mean and 95th percentile values of AUC of time-intensity curve were calculated at 60, 90, and 120 s of Enhancement-cycle-1 and -2. Optimal DCE parameters were identified by ROC analysis. Diagnostic performance of conventional MRI and DCE-MRI was compared using McNemar’s test. Results On WE, residual tumor was present in 23 (71.9%) of 32 patients. On MRI, Lesion-type-1 was found in 16/32 (50%) patients and Lesion-type-2 in 16/32 (50%). The optimal DCE parameter was the 95th percentile value of AUC at 120s of Enhancement-cycle-2. The sensitivity, specificity, and AUC were as follows: 65.2% (95% CI, 45.8–85.7%), 88.9% (CI, 68.4–100%), and 0.77 (CI, 0.62–0.92) for conventional MRI, and 100%, 55.6% (CI, 23.1–88.0%), and 0.78 (CI, 0.61–0.95) for combined conventional and DCE-MRI. Conclusions Additional DCE-MRI aided in detecting residual STS after UPE, particularly in cases without definite soft tissue nodular enhancement. Advances in knowledge Close follow up may be suggested for patients showing abnormality in DCE-MRI, with more suspicion of residual tumor.