This narrative review systematically assesses the diagnostic performance of hematological biomarkers as non-invasive markers for early identification of pre-cancerous states. Employing a comprehensive search across PubMed, Scopus, and Web of Science, it examines how persistent inflammation in the premalignant microenvironment induces detectable systemic changes through standard blood tests, such as neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and red cell distribution width (RDW). In contrast to prior reviews emphasizing established cancers, this synthesis focuses on premalignant conditions like oral potentially malignant disorders (OPMDs), cervical intraepithelial neoplasia (CIN), and Barrett’s esophagus, incorporating available sensitivity, specificity, and AUC data. Although these markers are highly accessible, their specificity is frequently compromised by overlapping inflammatory conditions. Thus, combining complete blood count (CBC) elements into multi-biomarker panels or using dynamic tools like the Personalized Indicator of Thrombocytosis (PIT) for serial tracking yields better prognostic accuracy than single cutoffs. The review also explores enhancements via machine learning models and molecular add-ons, such as cell-free DNA (cfDNA), for improved risk assessment. While issues like assay variability and inconsistent reference ranges persist, these blood-based biomarkers offer an affordable, adaptable strategy for surveilling transformation risk. By tackling deployment challenges and common confounders, this work outlines a practical roadmap for leveraging routine hematological indices to boost early detection and patient outcomes in pre-cancerous cohorts, especially in low-resource settings.
Mandefro et al. (2026) studied this question.