M Reyes Oliver-Perez1, Javier Magriña2, Cecilia Villalain-Gonzalez3, Jesus S Jimenez-Lopez4, Gregorio Lopez-Gonzalez3, Carmen Barcena5, Concepcion Martinez-Biosques6, Blanca Gil-Ibañez3, Alvaro Tejerizo-Garcia3. 1. Department of Obstetrics and Gynecology. University Hospital 12 de Octubre. Madrid, Spain. Instituto de Investigacion Sanitaria Hospital 12 de Octubre (imas12), Madrid, Spain. Universidad Complutense de Madrid, Madrid, Spain. Electronic address: m.delosreyes.oliver@salud.madrid.org. 2. Department of Medical and Surgical Gynecology. Mayo Clinic. Phoenix, AZ, USA. 3. Department of Obstetrics and Gynecology. University Hospital 12 de Octubre. Madrid, Spain. Instituto de Investigacion Sanitaria Hospital 12 de Octubre (imas12), Madrid, Spain. Universidad Complutense de Madrid, Madrid, Spain. 4. Department of Obstetrics and Gynecology. Hospital Regional de Málaga, Andalucia, Spain. 5. Department of Pathology. University Hospital 12 de Octubre. Madrid, Spain. Instituto de Investigacion Sanitaria Hospital 12 de Octubre (imas12), Madrid, Spain. Universidad Complutense de Madrid, Madrid, Spain. 6. Department of Obstetrics and Gynecology. Hospital General de Villarobledo, Albacete, Spain.
Abstract
OBJECTIVE: To analyze histological factors possibly associated with lymphovascular space invasion (LVSI) and to determine which of those can act as independent surrogate markers. METHODS: Retrospective cohort study performed between January 2001 and December 2014. LVSI was defined as the presence of tumor cells inside a space completely surrounded by endothelial cells. Risk factors evaluated included myometrial invasion, tumor grade, size, location, and cervical invasion. Univariate logistical regression models were applied to study any possible association of LVSI with these factors. Values were adjusted by multivariate logistic regression analysis. RESULTS: A total of 327 patients with endometrial carcinoma treated in our Centre were included. LVSI was observed in 120 patients (36.7%). Lower uterine segment involvement (OR 5.21, 95% CI:2.6-10.4, p < 0.001) and size ≥2 cm (OR 2.62, 95% CI: 1.14-6.1, p < 0.001) were independent factors for LSVI in multivariate analysis. In univariate analysis, LVSI was a surrogate marker in type 1 tumors with deep myometrial invasion (IB, 51.9% vs. IA, 16.0%; p < 0.001), grade 3 (G3 55.8% vs. G1 16.2%; p < 0.001), size ≥2 cm (37.9% vs. 16.1%, p = 0.005), those with involving the lower segment of the uterus (58.9% vs. 22.5%, p < 0.001) and/or with cervical stromal invasion (65.4% vs. 26.1%, p < 0.001), and in type 2 tumors (61.5% vs. 30.5%, p < 0.001). The use of uterine manipulator did not increase the rate of LVSI (35.5% vs. 40.5%, p = 0.612) as compared to no manipulator use. CONCLUSIONS: Size ≥2 cm and involvement of the lower uterine segment are independent factors for LSVI, in type 1 tumors, which can be used for surgical planning. LVSI is also more common in type 1 tumors with deep myometrial invasion, grade 3 and/or cervical stromal invasion, and also in type 2 tumors. The use of a uterine manipulator does not increase LVSI.
OBJECTIVE: To analyze histological factors possibly associated with lymphovascular space invasion (LVSI) and to determine which of those can act as independent surrogate markers. METHODS: Retrospective cohort study performed between January 2001 and December 2014. LVSI was defined as the presence of tumor cells inside a space completely surrounded by endothelial cells. Risk factors evaluated included myometrial invasion, tumor grade, size, location, and cervical invasion. Univariate logistical regression models were applied to study any possible association of LVSI with these factors. Values were adjusted by multivariate logistic regression analysis. RESULTS: A total of 327 patients with endometrial carcinoma treated in our Centre were included. LVSI was observed in 120 patients (36.7%). Lower uterine segment involvement (OR 5.21, 95% CI:2.6-10.4, p < 0.001) and size ≥2 cm (OR 2.62, 95% CI: 1.14-6.1, p < 0.001) were independent factors for LSVI in multivariate analysis. In univariate analysis, LVSI was a surrogate marker in type 1 tumors with deep myometrial invasion (IB, 51.9% vs. IA, 16.0%; p < 0.001), grade 3 (G3 55.8% vs. G1 16.2%; p < 0.001), size ≥2 cm (37.9% vs. 16.1%, p = 0.005), those with involving the lower segment of the uterus (58.9% vs. 22.5%, p < 0.001) and/or with cervical stromal invasion (65.4% vs. 26.1%, p < 0.001), and in type 2 tumors (61.5% vs. 30.5%, p < 0.001). The use of uterine manipulator did not increase the rate of LVSI (35.5% vs. 40.5%, p = 0.612) as compared to no manipulator use. CONCLUSIONS: Size ≥2 cm and involvement of the lower uterine segment are independent factors for LSVI, in type 1 tumors, which can be used for surgical planning. LVSI is also more common in type 1 tumors with deep myometrial invasion, grade 3 and/or cervical stromal invasion, and also in type 2 tumors. The use of a uterine manipulator does not increase LVSI.