Predictors of Sentinel Lymph Node Metastasis in Postoperatively Upgraded Invasive Breast Carcinoma Patients

SIMPLE SUMMARY: Routine sentinel lymph node (SLN) biopsy (SLNB) is not necessary for breast-conserving surgery (BCS)-treated ductal carcinoma in situ (DCIS) but is indicated for underestimated invasive carcinoma (IC) postoperatively. In this retrospective study, we aimed to investigate the factors contributing to SLN metastasis in underestimated IC patients with an initial diagnosis of DCIS by core needle biopsy. We found that only the features related to IC, including an IC tumor size (>0.5 cm) and the presence of LVI, could be used as risk predictors of SLN metastasis. In the absence of any predictors, the incidence of positive SLNs was very low (2.6%) in the total population and extremely low (1.3%) in the BCS subgroup. Therefore, we suggest that SLNB may be omitted in patients who initially underwent BCS without risk predictors on final pathological assessment. However, further prospective studies are warranted before its clinical application. ABSTRACT: Sentinel lymph node (SLN) biopsy (SLNB) usually need not be simultaneously performed with breast-conserving surgery (BCS) for patients diagnosed with ductal carcinoma in situ (DCIS) by preoperative core needle biopsy (CNB), but must be performed once there is invasive carcinoma (IC) found postoperatively. This study aimed to investigate the factors contributing to SLN metastasis in underestimated IC patients with an initial diagnosis of DCIS by CNB. We retrospectively reviewed 1240 consecutive cases of DCIS by image-guided CNB from January 2010 to December 2017 and identified 316 underestimated IC cases with SLNB. Data on clinical characteristics, radiologic features, and final pathological findings were examined. Twenty-three patients (7.3%) had SLN metastasis. Multivariate analysis indicated that an IC tumor size > 0.5 cm (odds ratio: 3.11, p = 0.033) and the presence of lymphovascular invasion (odds ratio: 32.85, p < 0.0001) were independent risk predictors of SLN metastasis. In the absence of any predictors, the incidence of positive SLNs was very low (2.6%) in the total population and extremely low (1.3%) in the BCS subgroup. Therefore, omitting SLNB may be an acceptable option for patients who initially underwent BCS without risk predictors on final pathological assessment. Further prospective studies are necessary before clinical application.