Breast cancer (BC) is the most common cancer in women and the leading cause of cancer-related death worldwide. Systemic immune dysregulation is increasingly recognized as a critical component of BC pathophysiology. Therefore, the characterization of inflammatory mediators in this neoplasia has emerged as a focus of increasing research. In this study, we performed a multiplex analysis of 62 circulating immune-related molecules in serum samples from 229 BC patients prior to any treatment and 50 healthy controls. Serum samples were analyzed using the LEGENDplex™ protocol and measured by a spectral flow cytometer. Among the 62 circulating immune-related molecules initially analyzed, we identified a distinct serum signature comprising 27 soluble factors that significantly discriminated BC patients from healthy controls. The integration of these biomarkers into supervised machine learning models revealed that the Light Gradient Boosting Machine (LightGBM) algorithm achieved optimal diagnostic performance (AUC = 0.91; accuracy = 96.4%), with TSLP, IL-8, and IL-1β emerging as key predictive features. By integrating clinicopathological data, we identified significant associations between immune factor levels and tumor characteristics, including proliferation, lymph-node involvement, and histological grade. Stratification by molecular subtype also revealed distinct immunologic patterns. Significantly, elevated levels of IL-2RA and PD-1 were identified as promising candidates to predict shorter overall and disease-free survival, respectively. These findings provide novel insights into the systemic immune landscape of early-stage BC and support the use of circulating immune mediators as minimally invasive biomarkers for diagnosis and prognosis. However, external validation is needed to confirm the robustness and generalizability of these findings.
Montoyo-Pujol et al. (Thu,) studied this question.