Logistic Regression Model of Relationship between Breast Cancer Pathology Diagnosis with Metastasis

Abstract

Management of breast cancer requires a diagnosis of breast cancer, including an accurate pathology diagnosis to direct the provision of appropriate treatment. Part of inseparable diagnosis of breast cancer itself is determining the presence of metastasis. This study aims to determine the relationship model of pathological diagnosis with the possibility of metastasis to other organs of the breast. Pathology diagnosis results include tumor position, topography, behavior, grade, and tumor status. This study used secondary data from the Hospital Information System (SIRS) of the hospital in Makassar, Indonesia. Data obtained from the results of pathology and clinical examinations of breast tumor as many as 204 patients were hospitalized during the year of 2018. Binary regression analysis modeling was performed to see the relationship between types of pathological diagnosis (laterality, topography, behavior, grade, tumor status) of breast cancer with the possibility of metastases to other organs. The results of the analysis found a significant relationship between the position and status of the tumor with the presence of metastasis. The results of the metastatic binary logistic regression model was L(Y) = 1.330 – 0.670lat1 – 0.779lat2 – 0.724status Chi-square analysis between metastasis with surgery and chemotherapy found the significant relationships between metastasis with surgery (X2 = 15.951; p value = 0.001) and metastasis with chemotherapy (X2 = 20,796; p value = 0.001). The results of this analysis indicate that the position and status of the tumor have a relationship with the likelihood of metastasis. In addition, this metastasis further determines the management of treatment whether surgery or chemotherapy or both.