Abstract
Gynecologic oncologists have sought to define adequate lymphadenectomy. The purpose of this study is to determine the probability of detecting lymph node metastasis by lymph node count compared to number of nodal stations sampled.
This is a clinicopathologic review of surgically staged endometrial carcinoma patients from 2000 to 2008. Information was extracted from patients’ medical records. Student
t-test, Wilcoxon rank sum test, Chi-square and Fisher exact tests were used. Elimination logistic regression was performed to identify independent significant predictors of lymph node metastasis.
p
<
.05 was considered significant for all tests.
The study population consisted of 352 patients with a mean age of 65. Forty patients (11.36%) had lymph node metastasis. Number of nodes sampled was not associated with lymph node status on univariate analyses. Patients with lymph node metastases detected was increased when 8 or more nodal stations were sampled compared to less than 8 (19.4% vs. 9.8%,
p
=
.04). More significance was seen when 9 or more stations were sampled (32% vs. 9.8%,
p
=
.004). Multivariate logistic regression analysis, controlling for age, grade, depth of myometrial invasion, number of nodes sampled, and number of nodal stations sampled, found only grade (
p
=
.002), depth of myometrial invasion (
p
<
.0003), and sampling of 9 or more nodal stations (
p
=
.03) to be independent predictors of node status.
Lymph node count did not accurately predict risk of lymph node metastasis. Number of nodal stations sampled was a more precise predictor of lymph node metastases.
