Author's reply to: What is the prognostic importance of lymphovascular space invasion in the absence of lymph node metastasis for early-stage endometrial cancer?

Dear Editor,

We have sincerely appreciated the interest shown in our study [1] and the constructive comments by Aslan and Meydanli [2], and we would like to clarify and discuss their observations.

Authors affirm that presence of lymph-vascular space invasion (LVSI) constitutes a substantial risk for occult lymph node metastasis. This is a well-known notion and, as stated in the introduction of our paper, many series in the literature show an association between LVSI and nodal metastasis in International Federation of Gynaecology and Obstetrics stage I–III patients. However, our aim was to evaluate in selective way patients with low-risk features. Certainly, almost half of our patients did not undergo nodal staging but in the era before sentinel lymph node biopsy, European guidelines (ESMO-ESGO-ESTRO consensus conference on endometrial cancer, 2015) considered lymphadenectomy not recommended for patients with low-risk endometrioid carcinoma (grade 1 or 2 and superficial myometrial invasion <50%) due to a low risk of lymph node involvement, and the absence of survival benefit proven in 2 randomized controlled trials.

Therefore, among 228 patients not staged, 22 had LVSI, 16 focal and 6 diffuse. Among the 6 patients with diffuse LVSI none was submitted to adjuvant treatment and we did not report any lymph nodal recurrence. We believe that this data could be a surrogate of absence of occult nodal metastasis, especially with no adjuvant treatment administered. Only 1 patient had a recurrence 6 months after surgery, in the pelvis and at the level of the soma D11.

More in general, considering all the 228 not completely staged patients, regardless LVSI status, we reported 6 recurrences. Only 1 patient had lymph nodal recurrence, but it occurred 72 months after surgery and the patient was not submitted to any adjuvant treatment.

For these reasons we are assuming that our population is generally homogenous, despite the bias of the absence of nodal staging, and the risk of occult nodal metastasis is very low.

Moreover, if we include in the analysis only the 296 completely staged patients, our results are still maintained. The risk of distant metastasis is 2.3% among patients with no LVSI, versus 25% among patients with diffuse LVSI (p=0.002). Diffuse LVSI was a prognostic factor for both DFS (hazard ratio [HR]=7.8; 95% confidence interval [CI]=2.78–21.88; p<0.001) and OS (HR=22.6; 95% CI=4.46–114.88; p<0.001).