Abstract P2-04-18: Unlocking the Complete Blood Count as a Risk Stratification Tool for Breast Cancer Using Machine Learning: A Large Scale Retrospective Study

Abstract Optimizing early breast cancer (BC) detection requires effective risk assessment tools. This retrospective study from Brazil showcases the efficacy of machine learning in discerning complex patterns within routine blood tests, presenting a globally accessible and cost-effective approach for risk evaluation. We analyzed complete blood count (CBC) tests from 396,848 women aged 40-70, who underwent breast imaging or biopsies within six months after their CBC test. Of these, 2,861 (0.72%) were identified as cases: 1,882 with BC confirmed by anatomopathological tests, and 979 with highly suspicious imaging (BI-RADS 5). The remaining 393,987 participants (99.28%) were classified as controls, with BI-RADS 1 or 2 results. The database was divided into modeling (including training and validation) and testing sets based on diagnostic certainty. The testing set comprised cases confirmed by anatomopathology and controls cancer-free for 4.5 to 6.5 years post-CBC. While our dataset lacked specific indicators for identifying metastasis and lymph node status required for TNM staging, the BC study population in the testing set primarily consisted of early-stage carcinomas, as reflected by the fact that 24% of the identified carcinomas were in situ and 95% of the recorded invasive T stages were at T1-2. Statistical analysis revealed that age, hematocrit, hemoglobin, mean corpuscular hemoglobin (MCH), mean corpuscular volume (MCV), neutrophils-lymphocyte ratio (NLR), derived neutrophils-lymphocyte ratio (dNLR), platelet-lymphocyte ratio (PLR), systemic immune-inflammation index (SII), systemic inflammation response index (SIRI), and aggregate index of systemic inflammation (AISI) were significantly elevated in women with BC, whereas lymphocytes and the lymphocyte-monocyte ratio (LMR) were markedly lower. Given the statistical significance of all CBC-derived ratios, they present a valuable opportunity as cost-effective predictive biomarkers in BC. Furthermore, by harnessing Artificial Intelligence (AI), we developed a ridge regression model incorporating age, NLR, and RBC, achieving an AUC of 0.64 (95% CI 0.64-0.65), demonstrating superior performance over more complex models like LightGBM, while offering full interpretability. The explanatory analysis indicates that higher age and NLR, along with lower RBC levels, may increase BC risk. Age correlates with BC due to cumulative exposure, genetic mutations, and cellular aging. High NLR suggests inflammation, potentially impairing immune response and promoting tumor growth, while low RBC might indicate a chronic inflammatory state conducive to cancer progression. To emulate the use of a risk stratification tool, using the probabilistic output from this model, we divided the study population into four risk groups: high, moderate, average, and low risk, which obtained relative ratios of BC of 1.99, 1.32, 1.02, and 0.42, respectively. The aim of this stratification was to streamline prioritization, potentially improving the early detection of breast cancer, particularly in resource-limited environments. As a risk stratification tool, this model offers the potential for personalized breast cancer screening by prioritizing women based on their individual risk, thereby indicating a shift from a broad population strategy. To support its clinical application, external validation with diverse populations is of paramount importance. Citation Format: Pedro Henrique Souza, Daniella Castro Araujo, Bruno Aragão Rocha, Karina Braga Gomes, Daniel Noce da Silva, Vinicius Moura Ribeiro, Marco Aurelio Kohara, Fernanda Marana, Renata Andrade Bitar, Adriano Alonso Veloso, Maria Carolina Tostes Pintao, Flavia Helena da Silva, Celso Ferraz Viana. Unlocking the Complete Blood Count as a Risk Stratification Tool for Breast Cancer Using Machine Learning: A Large Scale Retrospective Study [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P2-04-18.