Introduction/Background The Uterine mass Magna Graecia (U.M.G.) risk index, resulting from the inverse relationship between LDH1 and LDH3, help clinicians in discriminating between no-risk and high-risk uterine masses. The aim of the present study was to verify whether other LDH isoenzymes interact with the U.M.G. index in better stratifying the risk of uterine sarcoma.

Methodology The U.M.G. database, (data from 2254 patients, 2211 uterine fibroids and 43 sarcomas) was assessed again. A detailed exploratory analysis was performed and a machine learning technique was employed for identifying which were the most accurate indicators to classify, in association with the U.M.G. risk index, the risk of malignancy among uterine masses.

Results Tree indicators of sarcoma risk were identified: total LDH, LDH5 and point ‘p’ [p(LDH5, UMG)] (figure 1). Table 1 shows cut-off values for each indicator. UMG risk index, total LDH, LDH5 and point ‘p’, were integrated into an algorithm for the Risk Assessment in Uterine Lesions (R.A.U.L.). that allows to classify our population of women, with an accuracy closed to 100%, into 3 classes of risk: class A (no-risk), B (low-risk) and C (high-risk). When two or three indicators are in ‘class c’ there is a high risk of sarcoma; when three indicators are in ‘class a’ there is no risk of sarcoma; when indicators do not fall into the above two conditions, a low risk of sarcoma has to be considered ‘class b’.

• Download figure
• Open in new tab
• Download powerpoint

Abstract 2022-RA-1410-ESGO Figure 1

Conclusion An accurate risk assessment in uterine lesions would suggest clinicians which is the most appropriate diagnostic and therapeutic approach for each affected woman.The new patented algorithm R.A.U.L., once validated by prospective studies, would allow to better stratify the risk of sarcoma in order to limit open approaches and offer conservative treatment in women with no or low-risk and ensure oncological safe procedures in women at high-risk.