Preoperative diagnosis of uterine sarcoma is very difficult, and currently, its diagnostic accuracy is not satisfactory. It is therefore important to perform surgery and establish the pathological diagnosis if the clinical findings and various examination findings indicate possible uterine sarcoma. We investigated the accuracy of the combination of various types of predictors of uterine sarcoma and the novel PREoperative Sarcoma Score (PRESS) for avoiding unnecessary surgery while diagnosing uterine sarcoma. We retrospectively analyzed the clinical findings, blood tests, imaging studies (ultrasonography and magnetic resonance imaging [MRI]), and endometrial cytology of 63 suspected uterine sarcoma cases that underwent surgery from 2006 to 2012. These cases were also scored retrospectively using PRESS. We analyzed the number of unnecessary surgeries that could be avoided using PRESS. Of 63 cases, 15 were diagnosed with uterine sarcoma (sarcoma group), and 48 had benign tumors (benign group). Univariate analysis indicated age, serum lactate dehydrogenase (LDH) values, and MRI and endometrial cytology findings as significant predictors of uterine sarcoma in both groups. In contrast, multivariable analysis identified only age, serum LDH value, and endometrial cytology findings as significant predictors. Accordingly, the latter were placed as 2 points, and the remaining MRI finding as 1 point. The accuracy rate of prediction was 84.1%, and the positive and negative predictive values were 63.2% and 93.2% respectively when the PRESS was interpreted as "positive" when it was 3 points or higher. Using multiple predictors for the preoperative diagnosis of uterine sarcoma, our proposed PRESS score is beneficial in the clinical setting while making treatment decisions in suspected uterine sarcoma cases as well as avoiding unnecessary surgery.
Keywords: Multiple predictors; Preoperative diagnosis; Uterine sarcoma.