P04.01: Preoperative diagnosis of uterine sarcoma using tumour signal intensity at T2 weighted imaging and apparent diffusion coefficient (ADC) of magnetic resonance imaging (MRI)

Abstract

Uterine sarcoma is a rare and poor prognostic disease, and its pre-operative diagnosis from uterine tumour is difficult. We focused on the tumour-to-subcutaneous fat signal intensity ratio at MRI T2 weighted imaging for the pre-operative diagnosis of uterine sarcoma. We had already found that a tumour with a ratio of over 0.512 is highly suspected of sarcoma. However, although the sensitivity is 100%, specificity is 75% which is not high enough for clinical use. The purpose of this study is to make the preoperative diagnosis method using MRI more reliable through additional investigation. MRI was performed in 4 cases (5 tumours) of uterine sarcoma in the last 2 years at our hospital (sarcoma group). In contrast, 31 cases of leiomyoma received MRI in 2015 (myoma group). We calculated tumour-to-subcutaneous fat signal intensity ratios on MRI T2 weighted imaging and ADC for both groups and made comparisons. The cut-off value of ADC to predict sarcoma was also calculated. Finally, we considered a more reliable way of preoperative diagnosis for uterine sarcoma by using a combination of these parameters. The sarcoma group had significantly lower ADC when compared with the leiomyoma group (1.10 ± 0.251 vs. 1.45 ± 0.194, p < 0.01). The best cut-off value of ADC to diagnose sarcoma is under 1.34 (sensitivity: 100%, specificity: 84%, area under the curve: 0.950). All sarcoma showed over 0.512 tumour-to-subcutaneous fat signal intensity ratios and under 1.34 ADC. On the other hand, no leiomyomas showed ratios over 0.512 together with ADC under 1.34. The uterine tumours which showed over 0.512 tumour-to-subcutaneous fat signal intensity ratios on T2 weighted imaging and under 1.34 ADC are likely to be sarcoma. All uterine tumours other than these tumours are likely to be leiomyoma. This suggest that pre-operative diagnosis of uterine sarcoma is possible by a combined use of tumour-to-subcutaneous fat signal intensity ratio on MRI T2 weighted imaging and ADC at MRI.