Blood-based biomarkers for breast cancer screening and early diagnosis: Diagnostic performance and challenges for clinical implementation.

The diagnosis of breast cancer (BC), especially early-stage BC, remains a significant clinical challenge due to the limited sensitivity and specificity of conventional imaging and serum tumor markers for screening. In this review, we examine the current landscape of blood-based biomarkers for BC diagnosis, including circulating tumor DNA (ctDNA), microRNAs (miRNAs), extracellular vesicles (EVs), autoantibodies, and proteomic/metabolomic signatures, highlighting their diagnostic performance and their limitations. There are dozens of published works reporting promising accuracy for BC detection using many of these biomarkers, often with areas under the receiver operating characteristic (ROC) curve exceeding 0.90 in selected cohorts. Yet, their clinical implementation is impaired by numerous methodological obstacles (sparse abundance of ctDNA, variability of methodologies of detection of miRNAs, difficulty of isolation of EVs, or incomparability of platforms of proteomics and metabolomics), and of study-design (uni- vs multicenter study, selection of control groups, underrepresentation of some ethnic populations) nature. Rather than just providing a comprehensive catalog, we focus on the unique unsolved challenges of each biomarker class and discuss strategies to overcome them. Future progress will depend on the standardization of methodologies, validation in large multicenter cohorts, and integration of multi-omic biomarker panels supported by advanced computational approaches. Describing both the prospects and lingering obstacles of their realization allows this work to propose a roadmap for the development of clinically valid blood-based biomarkers for the early detection of breast cancer.