PURPOSE: Evaluation of lymph node (LN) metastases in endometrial carcinoma (EC) is an important prognostic factor and a required element of cancer staging. The purpose of this study is to analyze what factors might predict the likelihood of nodal involvement in EC. METHODS: A retrospective search of our institutional database for hysterectomies with associated LN dissection in women with EC revealed 207 cases between 2005 and 2012. Cases with primary EC, irrespective of histologic subtype, including carcinosarcomas were included in the study, but pure sarcomas were excluded. We evaluated various factors including tumor size (TS; ≤2.0 cm and >2.0 cm), depth of myometrial invasion (DMI; absent, ≤50 %, >50 %), positive pelvic cytology (PPC), cervical stromal invasion (CSI), and lymph-vascular invasion (LVI), to determine which factors correlated with the presence of LN metastasis. RESULTS: Of the 207 (age = 62.29 ± 10.9, mean ± SD) cases of EC with LN dissection in our study group, 34 (16.42 %) had positive LNs. On univariate analysis, we found that TS (p = 0.04), tumor grade (Grade I and II versus grade III, p < 0.0001), DMI (p < 0.0001), CSI (p < 0.0001), LVI (p < 0.0001), and PPC (p = 0.001) showed statistically significant correlation with LN metastasis. However, on multivariate analysis, only DMI (p = 0.002) and LVI (p = 0.004) independently showed statistically significant correlation with LN metastasis. In addition, 18 (8.7 %) grade I and II (well/moderately differentiated) tumors with TS ≤2.0 cm and <50 % DMI showed no LN metastasis, LVI, CSI, or PPC. CONCLUSION: We concluded that DMI and LVI were independent factors predictive of LN metastasis.
PURPOSE: Evaluation of lymph node (LN) metastases in endometrial carcinoma (EC) is an important prognostic factor and a required element of cancer staging. The purpose of this study is to analyze what factors might predict the likelihood of nodal involvement in EC. METHODS: A retrospective search of our institutional database for hysterectomies with associated LN dissection in women with EC revealed 207 cases between 2005 and 2012. Cases with primary EC, irrespective of histologic subtype, including carcinosarcomas were included in the study, but pure sarcomas were excluded. We evaluated various factors including tumor size (TS; ≤2.0 cm and >2.0 cm), depth of myometrial invasion (DMI; absent, ≤50 %, >50 %), positive pelvic cytology (PPC), cervical stromal invasion (CSI), and lymph-vascular invasion (LVI), to determine which factors correlated with the presence of LN metastasis. RESULTS: Of the 207 (age = 62.29 ± 10.9, mean ± SD) cases of EC with LN dissection in our study group, 34 (16.42 %) had positive LNs. On univariate analysis, we found that TS (p = 0.04), tumor grade (Grade I and II versus grade III, p < 0.0001), DMI (p < 0.0001), CSI (p < 0.0001), LVI (p < 0.0001), and PPC (p = 0.001) showed statistically significant correlation with LN metastasis. However, on multivariate analysis, only DMI (p = 0.002) and LVI (p = 0.004) independently showed statistically significant correlation with LN metastasis. In addition, 18 (8.7 %) grade I and II (well/moderately differentiated) tumors with TS ≤2.0 cm and <50 % DMI showed no LN metastasis, LVI, CSI, or PPC. CONCLUSION: We concluded that DMI and LVI were independent factors predictive of LN metastasis.