Determining preoperative predicting value of adnexal masses combining clinical, ultrasonographic and blood parameters

Purpose: Purpose of this study is to create a reliable model for predicting pre operative malignity risk of patients with adnexal mass diagnosis combining clinical symptoms, radiologic appearance and laboratory findings. Methods: Preoperative clinical symptoms, ultrasound findings, tumor markers (CA 125, CA 19-9, CA 15-3, CEA, AFP), CBC parameters with neutrophil-lymphocyte ratio and platelet-lymphocyte ratio was analyzed in 636 patients undergoing surgery with a diagnosis of adnexal mass between the dates 2010-2016. Symptoms were divided into two groups (benign (n=509) and borderline/malign (n=127) ovary tumor) according to postoperative histopathologic examination. Examined parameters were divided into categorical values which showed a significant difference between the two groups and parameters affecting the preoperative prognosis of malignant masses were found according to logistic regression analysis. A risk likelihood score (ATATÜRK score model) was performed by giving scores to each parameter of the effectors of these factors, and cases were scored (minimum 1, maximum 7) according to this model. In order to evaluate the prediction performance of preoperative malignity of this model, positive predictive value, negative predictive value, positive likelihood ratio (L+) and negative likelihood ratio (L-) were calculated. Statistical analyses were performed using SPSS (version 17.0) and p value <0,05 was considered statistically significant. Results: Statistically significant parameters between two groups (benign and malign histopathology) were; age, gravida, parity, duration of menopause, CA 125, CA 15-3, CA 19-9, CEA, and erythrocyte count, hematocrit ratio, neutrophil count, neutrophil ratio, lymphocyte ratio, monocyte ratio, platelet count, PCT, neutrophil/lymphocyte ratio, platelet-lymphocyte ratio from hematologic parameters (p<0,05). According to the results of multivariant logistic regression analyze; postmenopausal period, ≥ 2 pathologic findings in ultrasonography, CA 125 value ≥ 45 U/ml, CA 15-3 value ≥20 U/ml and platelet-lymphocyte ratio ≥155 were significant factors in predicting malignancy risk of the mass. Prediction of malign masses pre operatively was calculated according to the scoring model with an accuracy rate of 79,5%, when total score ≥ 4, the sensitivity of 83,1%, specificity of 78,5%, the positive predictive value of 51,3%, the negative predictive value of 94,4%, L+=3,86, l-=0,22. Conclusion: Malignity of adnexal mass could be predicted by using recently constituted ATATÜRK scoring model with high sensitivity. This scoring model enables patients with high malignity risk could be dispatched to specialized gynecologic centers for attaining adequate surgical treatments by using this scoring model. Keywords: Adnexal mass, pelvic mass, ovarian mass, malignity, preoperative prediction