The use of ultrasound in the prediction of endometrial cancer in women with postmenopausal bleeding

Sammanfattning: The aim of the work presented in this thesis was to identify the ultrasound methods and ultrasound features that are most useful for the prediction of endometrial cancer in patients with postmenopausal bleeding and sonographic endometrial thickness 4.5 mm or more. The grey-scale ultrasound variable that best predicted malignancy was heterogeneous endometrial echogenicity (area under the receiver operating characteristics curve, AUC, 0.83), and the power Doppler ultrasound variable that best predicted malignancy was irregular branching of endometrial blood vessels (AUC 0.77). Mathematical models for evaluation of the individual risk of endometrial malignancy were constructed using clinical data, sonographic endometrial thickness, grey-scale ultrasound morphology of the endometrium, and power Doppler ultrasound findings. The model with the largest AUC (0.91) and the highest specificity at 90% sensitivity included the variables endometrial thickness, Vascularity index (reflecting the vascularization of the endometrium when using power Doppler ultrasound), age, and use of hormone replacement therapy. The endometrial volume as measured by three-dimensional (3D) ultrasound was larger and the 3D power Doppler flow indices (reflecting endometrial vascularization) were higher in both the endometrium and in the subendometrium in women with malignant endometrium than in those with benign endometrium, but there was substantial overlap between the two categories. The saline contrast sonohysterography (SCSH) variable that best discriminated between benignity and malignancy for both two-dimensional (2D) and 3D SCSH was the presence of at least one focal lesion with an irregular surface (for 2D SCSH the AUC was 0.84, for 3D SCSH the AUC was 0.70). The mathematical models to estimate the individual risk of endometrial malignancy constructed in this thesis are likely to be useful for individualizing the management of women with postmenopausal bleeding and sonographic endometrial thickness 4.5 mm or more, but they need to be prospectively validated before they can be introduced into clinical practice.