Malignancy Algorithms in The Preoperative Assessment of Suspicious Adnexal Masses During Pregnancy

Abstract

INTRODUCTION: The aim of the present study was to evaluate the effectiveness of the three different algorithms on pregnant women who had adnexal mass, and also, to compare the sensitivity and specificity of these algorithms in pregnancy. METHODS: Our retrospective study evaluated the women who had a suspicious adnexal mass during pregnancy consulted to the Division of Gynecologic Oncology and underwent surgery at the Department of Obstetrics and Gynecology, Istanbul University from December 1999 to December 2019. RESULTS: There were forty pregnant women with adnexal mass. Eleven (30%) has benign, six (15%) had borderline, and twenty-three (55%) had malign lesions. The RMI4 algorithm had a sensitivity of 0.21, a specificity of 0.91, at a cut-off point ? 450. At a cut-off>%10, the LR2 and ADNEX algorithms both had a sensitivity of 1.00, however, LR2 algorithm had a specificity of 0.55; ADNEX algorithm had a specificity of 0.82. Positive predictive values (PPV) were 0.86 for RMI4, 0.85 for LR2 and 0.94 for ADNEX algorithm, and negative predictive values (NPV) were 0.30 for RMI4, 1.00 for LR2 and ADNEX algorithms. The difference between RMI4 and both LR2 and ADNEX algorithm was statistically significant (p <0.001 for all). However, the difference between the LR2 and ADNEX algorithm was not significant (p=0.25). DISCUSSION AND CONCLUSION: The ADNEX and LR2 algorithms had high sensitivity in differentiating malignant tumors from benign lesions on pregnant women with adnexal mass. In addition, ADNEX had the highest specificity of all. However, RMI4 algorithm had the poorest performance on pregnant women with adnexal mass.