Estudio de los factores predictivos del resultado del análisis intraoperatorio de la biopsia selectiva del ganglio centinela en el cáncer de mama y su implicación en la indicación del mismo

Abstract

ABSTRACT In the management of breast cancer, adequate staging is essential, including knowledge of the nodal involvement of the cancer. In this context, there are antecedents that imply a more conservative management of the axilla from the surgical point of view, avoiding lymphadenectomies through the use of selective sentinel lymph node biopsy (SLNB) when identifying nodes with or without involvement. Historically, there has been an increasingly conservative approach, depending on the intensity of such involvement, moving from the requirement of negativity, to consider the presence of isolated tumor cells and micrometastases, and even, in the last two decades, the presence of node involvement in the sentinel with macrometastases. All of the above highlights the usefulness of being able to know, at the time of surgery, the probability of involvement in order to decide whether or not to proceed with lymphadenectomy. However, this evolution to an increasingly conservative attitude, together with the necessary investment of resources in time and diagnostic methods that intraoperative analysis of the sentinel node entails, leads to the consideration of the convenience of knowing, as accurately as possible, the existence of variables that predict the conditions of nodal involvement and can avoid the need for intraoperative analysis. This project proposes an analysis of preoperative variables that allow predicting with sufficient precision the tumor load in the sentinel node or nodes (SNs) that require lymphadenectomy to be performed in the same surgical act. For this purpose, a retrospective study of cases with invasive breast cancer in which the SLNB was performed, collecting variables (clinical, imaging method, histopathological, immunohistochemical), obtained prior to surgery, and conducting a study of possible relationship with a tumor load and current criteria that require lymphadenectomy. In a global way, in the bivariate analysis, the variables that were related in a statistically significant way with dependent variable "LA indication after SN IO analysis result" and/or with "SN IO affectation" were: menopause, tumor size, palpability, number of lesions, number of isolated SNs, lymphovascular invasion, differentiation degree, percentage of ductal carcinoma in situ, and subrogate molecular classification. In the multivariate logistic regression analysis, lymphovascular invasion, differentiation degree, number of isolated SNs, and subrogated molecular classification were kept statistically significant. Three mathematical models were obtained, one for the H&E technique, another for the OSNA method and another for the general series, which would allow identifying low risk groups in which to avoid the intraoperative analysis of the sentinel node. As a conclusion, the presence of lymphovascular invasion and a low degree of histological differentiation in the preoperative biopsy; a higher number of two isolated SNs; and, when considering the molecular method with OSNA, the immunohistochemical profile, has relation with the probability that the intraoperative analysis of the SN determines a change of attitude in the management of the axilla during the same surgical act.