Many studies on the pre-operative prediction of myo- metrial invasion in patients with endometrial cancer used different morphological and color Doppler parameters with varying results. We tested their individual value and constructed a new logistic regression model.
Data from 104 consecutive patients with endometrial cancer were prospectively collected. All patients underwent pre-operative ultrasound examination with color Doppler imaging by the same gynecologist.
Ninety-seven women underwent full surgical staging. Their mean age was 65.9 years and 88 patients were postmenopausal. Fifty- nine had FIGO stage Ia or Ib disease, whereas 38 had deep myometrial invasion or higher stage disease. Because of missing values the coefficients of the logistic regression model were based on 94 patients. The results are summarized in the table. Multivariate logistic regression analysis retained 4 variables: the endometrial thickness and volume, the degree of differentiation, and the number of fibroids.
Single morphological or color Doppler ultrasound parameters cannot reliably predict the depth of myometrial invasion. Moreover, the Doppler ultrasound parameters do not contribute in this prediction. The logistic regression model was significantly better (p = 0.03, one-sided test) than the subjective assessment of the gynecologist. Prospective validation studies are needed to test its usefulness in clinical practice.
Unlabelled TableThreshold Area underROC Sensitivity(%) Specificity(%) Accuracy(%)Myometrial thickness8 mm0.71746166Endometrial thickness (ET)14 mm0.76816471ET/A-P diameter0.4290.75727172Endometrial volume (EV)4.93 ml0.76716970EV/uterine volume0.0850.78698076Subjective assessment-0.79616876Logistic regression model0.450.89778683

International journal of gynecological cancer, 2003-03, Vol.13 (Suppl 1), p.26-26

Pre-Operative Prediction of Depth of Myometrial Invasion in Patients With Endometrial Cancer: Evaluation of Ultrasound Parameters and Development of a New Logistic Regression Model