The Reliability of Intraoperative Assessment on Predicting Tumor Size, Myometrial Invasion, and Cervical Involvement in Patients With a Preoperative Diagnosis of Complex Atypical Hyperplasia or (Clinical Stage I) Endometrial Cancer: A Prospective Cohort Study.

ABSTRACT

OBJECTIVES: The objective of this study is to assess the reliability of intraoperative uterine assessment compared with the final pathologic evaluation in patients with endometrial cancer (EC) and whether assessment improves with experience. METHODS: After Institutional Review Board approval, a prospective cohort study of women surgically managed with biopsy-proven complex atypical hyperplasia (CAH) or EC between March 2015 and December 2016 was performed. Demographics, preoperative biopsy results, procedure, intraoperative and final pathologic evaluation of lesion size, myometrial invasion, and lower uterine segment/cervical involvement were abstracted. The agreement between the intraoperative and final pathologic evaluation of tumor involvement of the uterus was determined using the kappa statistic and the intraclass correlation coefficient. RESULTS: A total of 264 patients with a preoperative diagnosis of CAH or EC were included-71 (26.9%) with CAH and 193 (73.1%) with EC. The mean age was 62.6±11.5, and mean body mass index was 37.2±10.1. The majority of women were white (67%). A total of 227 (85.9%) patients underwent a laparoscopic or robotic hysterectomy, whereas 36 (13.6%) underwent an abdominal hysterectomy. 233 (88.3%) patients had EC and 21 (7.9%) patients had CAH on final pathology. There was a fair agreement between the intraoperative estimation of myometrial invasion (κ=0.37). A moderate agreement exists between the intraoperative estimation of lower uterine segment/cervical involvement (κ=0.57). There was a strong agreement between intraoperative tumor size assessment and the final path (intraclass correlation coefficient=0.74). The intraoperative correlation of tumor size was similar for the first half of the cohort (κ=0.50) and the second half (κ=0.46) chronologically. CONCLUSIONS: Despite only a fair correlation in the myometrial invasion, intraoperative assessment of cervical involvement and especially tumor size is more readily identified and overall accurate. Therefore, intraoperative evaluation is an additional tool to use when making the decision to proceed with surgical staging.