OBJECTIVES: To evaluate the accuracy of visual examination of myometrial invasion and the involvement of the cervix in the hysterectomy specimen and to explore the role of tumor grade and size in the accuracy of gross estimation of myometrial invasion. METHODS: In 142 patients with apparent early endometrial cancer the uterus was opened after its removal and inspected. The size of the tumor (<or= or >2 cm), the depth of myometrial invasion (less or greater than 50%) and the involvement of the cervix (as positive or negative) were visually estimated and recorded. All patients underwent surgical staging. The gross findings were compared with the final histological results. Estimations of myometrial invasiveness were analyzed according to the tumor grade and size. Accuracy, specificity, sensitivity, positive and negative predictive values were calculated. RESULTS: The overall accuracy rate for myometrial invasion was 81.7% (116/142). False positive and false negative results noted in 17/101 (17%) and 9/41 (21.9%) of patients. Sensitivity, specificity, positive and negative predictive values were 78%, 83.2%, 65.3% and 90.3% respectively. Gross estimation of invasion was more often successful in patients with smaller (<or=2 cm) than in those with grater tumors (accuracy 88.9% vs. 79.2%). Increasing tumor grade found to diminish the rate of correct prediction. The accuracy for grade I tumors was 93.5%, for grade II 80.4% and only 58.6% for grade III lesions. Cervical involvement was correctly evaluated in 138/142 patients with one false positive and 3 false negative results. CONCLUSION: These data show that gross estimation of myometrial invasion is highly accurate in small (<or=2 cm) and grade I tumors. Visual evaluation of cervical involvement is also reliable. However, prediction of myometrial invasion is lower in tumors with higher grade, being poor in grade III lesions. Inaccuracies of preoperative histology may decrease the sensitivity of intraoperative assessment based on final grade.
OBJECTIVES: To evaluate the accuracy of visual examination of myometrial invasion and the involvement of the cervix in the hysterectomy specimen and to explore the role of tumor grade and size in the accuracy of gross estimation of myometrial invasion. METHODS: In 142 patients with apparent early endometrial cancer the uterus was opened after its removal and inspected. The size of the tumor (<or= or >2 cm), the depth of myometrial invasion (less or greater than 50%) and the involvement of the cervix (as positive or negative) were visually estimated and recorded. All patients underwent surgical staging. The gross findings were compared with the final histological results. Estimations of myometrial invasiveness were analyzed according to the tumor grade and size. Accuracy, specificity, sensitivity, positive and negative predictive values were calculated. RESULTS: The overall accuracy rate for myometrial invasion was 81.7% (116/142). False positive and false negative results noted in 17/101 (17%) and 9/41 (21.9%) of patients. Sensitivity, specificity, positive and negative predictive values were 78%, 83.2%, 65.3% and 90.3% respectively. Gross estimation of invasion was more often successful in patients with smaller (<or=2 cm) than in those with grater tumors (accuracy 88.9% vs. 79.2%). Increasing tumor grade found to diminish the rate of correct prediction. The accuracy for grade I tumors was 93.5%, for grade II 80.4% and only 58.6% for grade III lesions. Cervical involvement was correctly evaluated in 138/142 patients with one false positive and 3 false negative results. CONCLUSION: These data show that gross estimation of myometrial invasion is highly accurate in small (<or=2 cm) and grade I tumors. Visual evaluation of cervical involvement is also reliable. However, prediction of myometrial invasion is lower in tumors with higher grade, being poor in grade III lesions. Inaccuracies of preoperative histology may decrease the sensitivity of intraoperative assessment based on final grade.