Can clinicopathological factors improve the prediction of metastasis to nonsentinel lymph nodes in patients with breast cancer?

Abstract

To determine whether clinicopathological characteristics can improve the prediction of metastasis to nonsentinel lymph nodes (NSLNs) over the use of only mRNA copy number in sentinel lymph node (SLN) biopsies. This was a retrospective, observational study that included a total of 824 patients with T1-3 breast cancer who had clinically negative, ultrasound-negative axilla without evidence of metastasis and who underwent one-step nucleic acid amplification in SLN biopsies. 118 required a complete axillary lymph node dissection (ALNhD). About 35.6% (42/118) had metastases to a NSLN, and 64.4% (76/118) had no metastasis to a NSLN. The ROC curve of the total tumor load (TTL) presented an area under the curve (AUC) of 0.651 (95%; CI: 0.552-0.751). The 7294 copies of CK19 mRNA were established as the optimal cutoff point, with sensitivity: 93%, specificity: 63%, positive predictive value: 44%, and negative predictive value: 91%. By associating the clinicopathological parameters (multicentricity, pooled immunohistochemistry [IHC], and progesterone receptors), the AUC went up to 0.752 (95% CI: 0.663-0.841). Clinicopathological factors should be considered together with the total CK19 mRNA copy number (the TTL) of the SLNs to improve the predictive capacity of metastatic involvement of the NSLNs.