IMR Press / EJGO / Volume 41 / Issue 5 / DOI: 10.31083/j.ejgo.2020.05.5401
Open Access Original Research
Uterine Cancer Normogram to Predict Lymph Node Metastasis: Comparison to the Mayo Algorithm and an External Validation of a Model in a North American Population
Show Less
1 Gynecologic Oncology, Kaiser Permanente Washington, United States
Eur. J. Gynaecol. Oncol. 2020 , 41(5), 681–684;
Submitted: 24 October 2019 | Accepted: 17 April 2020 | Published: 15 October 2020

Objective: We sought to compare two intraoperative uterine cancer normograms for prediction of lymph node (LN) metastasis. We used the widely known Mayo criteria, comparing it to an algorithm provided by Koskas et al. to predict likelihood of LN metastasis. Design: 490 uterine cancer patients from a single practice provider were included in the review. Data was abstracted to include age, race, stage, tumor size, grade, histologic subtype, depth of invasion, cervical involvement, lymphovascular space involvement (LVSI), and microsatellite instability (MSI). Patient comorbidities were analyzed to include body mass index (BMI), diabetes, and hypertension. Laboratories for these comorbidities were included. Those patients staged 1, 2, and 3 were included in final analysis. Results: The receiver operator curve (ROC) for the Koskas normogram was 0.78 when 4% was used as the cutoff for LN metastasis, with a sensitivity of 78% and specificity of 60%. When a 5% cutoff was used, the ROC was 0.71. For every percentage point that the French score rose, the chance of being LN positive increased by 0.8% (p < 0.001). The three point Mayo criteria odds ratio (OR) was 7.4 and the ROC was 0.57. Lymph node metastasis also correlated with MSI as seen on immunohistochemistry (IHC) testing. Conclusions: The Koskas normogram provided a better predictive algorithm for risk assessment of LN metastasis. Our results are comparable with those previously published by Koskas et al. providing an external validation of this normogram previously used in an European population. These intraoperative variables can be incorporated into real time risk assessment for LN metastasis and operative decision making. Mayo criteria, not using tumor size, could spare an additional 40% of patients an unnecessary LND compared to standard 3 point Mayo criteria--with better predictive value.

Uterine cancer
Endometrial cancer
Lymph node metastasis
Figure 1.
Back to top