Accuracy of preoperative endometrial sampling diagnosis of FIGO grade 1 endometrial adenocarcinoma.
Academic Article
Overview
abstract
OBJECTIVE: To evaluate the ability of a preoperative diagnosis of FIGO grade 1 endometrial adenocarcinoma and intraoperative depth of myoinvasion (DOI) to predict low-risk (LR) and high-risk (HR) final uterine pathology. METHODS: We reviewed 1423 consecutive cases of endometrial cancer treated at our institution between 1/1/93 and 5/31/06 to identify cases with a preoperative endometrial biopsy demonstrating FIGO grade 1 endometrial adenocarcinoma. All cases were pathologically reviewed at our institution and underwent surgical therapy at our institution. We excluded equivocal preoperative biopsies as well as those with serous or clear cell histology. Final uterine pathologic findings were grouped into low- and high-risk. Chi-square and Fisher-exact tests were used as appropriate. RESULTS: We identified 490 cases with a median age of 60 years (range 29-90 years). In 482 cases in which final pathologic grade was assessable, FIGO grade was greater in 71 (14.7%) cases; (66 [13.7%] were grade 2, and 5 [1%] were grades 2-3/3). Serous or clear cell histology was diagnosed in 6 (1.2%) additional cases. HR final uterine pathology was seen in 86 (18.5%) cases. Frozen section assessment of DOI, when performed, was associated with HR pathology (p<0.001). HR pathology was present in 3 (3.6%) of 84 cases with either no tumor or myoinvasion identified on frozen section. Lymph node metastasis was identified in 9 (4.4%) of 205 patients that underwent nodal evaluation. CONCLUSIONS: Preoperative FIGO grade 1 diagnosis correlates with final post-hysterectomy grade in 85% of cases. The rate of HR uterine pathology based on preoperative grade 1 alone is 18.5%. Frozen section may help further stratify for the risk of final HR uterine pathology but is not entirely accurate. The rate of HR uterine pathology is 4% if no cancer or myoinvasion is identified on frozen section and 18% if myoinvasion up to 50% is identified.
