RESUMO
OBJECTIVE: To evaluate the accuracy of transvaginal ultrasound for the evaluation of myometrial invasion in endometrial carcinoma in comparison with standard paraffin section. METHOD: A total of 111 patients with endometrial carcinoma diagnosed from fractional curettage underwent pre-operative transvaginal ultrasonography to assess myometrial invasion. Operation for surgical staging was subsequently performed and the hysterectomy specimen was evaluated for depth of myometrial invasion by standard paraffin section blinded from transvaginal ultrasound results. Final histopathologic diagnosis and depth of myometrial invasion were obtained from standard paraffin section. Ultrasonographic assessment was compared with the histopathological results. RESULTS: In evaluation of myometrial invasion, transvaginal ultrasound yielded the sensitivity of 69.4 per cent, specificity of 70.6 per cent, positive predictive value of 53.2 per cent, negative predictive value of 82.8 per cent, and accuracy of 70.3 per cent. The accuracy, sensitivity, specificity, negative predictive value, false positive and negative rates were comparable between grade 3 and grade 1 and 2 tumors. However, the positive predictive value was significantly higher among grade 3 than grade 1 and 2 tumors. The Kappa coefficients were 0.57 and 0.22 for grade 3 and grade 1 and 2 tumors respectively. CONCLUSION: Transvaginal ultrasound for assessment of depth of myometrial invasion in endometrial carcinoma provided acceptable accuracy compared with standard paraffin section. This technique might be of value for the decision making in the intra-operative management of endometrial carcinoma.