Taxonomic Profiling of Systemic Inflammatory Parameters as Predictors of Tumor Progression in Primary Colorectal Cancer.

Colorectal cancer (CRC) is one of the most common malignancies worldwide, with systemic inflammation increasingly recognised as a determinant of disease progression. This study aimed to establish a taxonomy-based classification of patients with newly diagnosed primary CRC using systemic inflammatory, haematological, and anthropometric markers, and to evaluate its association with tumour stage. A total of 229 patients (111 women, 118 men) undergoing surgery for primary CRC were included. Blood samples were analysed for haemoglobin, leukocytes, neutrophils, lymphocytes, platelets, C-reactive protein (CRP), and carcinoembryonic antigen (CEA). Anthropometric data were collected. Taxonomic clustering and ordinal logistic regression were used to explore associations with TNM and Astler-Coller classifications. Men had higher neutrophil and leukocyte counts, elevated CEA concentrations (132.8 vs. 81.3 ng/mL), and higher NLR values (4.74 vs. 4.23) compared with women. Logistic regression confirmed that platelet count (OR 1.003; = 0.004), PLR (OR 1.003; = 0.003), and CEA (OR 1.03; < 0.001) were positively associated with advanced TNM stage, while haemoglobin was inversely correlated (OR 0.88; = 0.045). Among 84 clustering models, two taxonomies were the most clinically informative: Taxonomy I (BMI, neutrophils, platelets) and Taxonomy II (age, lymphocytes, platelets), both significantly associated with T, N, M, overall TNM stage, and Astler-Coller grade. Taxonomy I identified three patient groups. Type 3 represented the poorest phenotype, characterised by low BMI and haemoglobin, high platelets, elevated CEA and PLR, and predominance of TNM IIIC tumours, consistent with a cachectic-inflammatory profile. Type 1 displayed higher BMI, lower inflammation, and earlier-stage disease. Type 2 was characterized by elevated neutrophils and leukocytes. Taxonomy II distinguished four groups, with Type 2 demonstrating the most favourable profile (high haemoglobin and lymphocytes, low NLR and PLR, early TNM stage). Systemic inflammatory markers, haemoglobin, platelets, and CEA strongly predict CRC advancement. The proposed taxonomy provides clinically meaningful stratification of CRC patients and may support personalised risk assessment. This accessible approach may facilitate early identification of high-risk individuals, although validation in prospective studies is required.