Tobacco Aerosols, Lung Injury, and Cancer: Pathology, Precision Therapy, and AI Diagnostics

Abstract

Tobacco aerosols causes overlapping but unequal pulmonary injury. This review compares combustible cigarettes, electronic nicotine delivery systems (ENDS), and heated tobacco products (HTPs), while examining t inflammatory susceptibility of the host, ferroptosis-immune crosstalk, precision oncology, and artificial intelligence (AI). A structured narrative search of PubMed/MEDLINE, ClinicalTrials.gov, major guidelines, and authoritative regulatory sources prioritised 2015–2026 evidence and pivotal earlier studies. Evidence was graded by design, independence, clinical relevance, and traceability. Combustible smoking has definitive causal links to chronic obstructive pulmonary disease and lung cancer. ENDS are established causes of e-cigarette or vaping product use-associated lung injury, whereas long-term cancer risk remains unquantified. HTPs reduce selected emissions but retain toxicants, with no demonstrated reduction in cancer incidence. Candidate IL-10, IL-18, and related variants suggest inherited modulation of inflammatory susceptibility, but current associations are population-specific and not ready for clinical risk prediction. Ferroptosis interacts bidirectionally with CD8+ T cells and myeloid populations, making cell type and disease stage critical. AI may support screening and pathology triage, but heterogeneity, bias, limited external validation, explainability, workflow integration, and model drift constrain implementation. These findings support an exposure hierarchy rather than equivalence: shared injury pathways do not imply equal risk, and host-genetic, ferroptosis, and AI findings remain insufficiently validated for routine clinical use. Clinical translation should prioritise complete cessation, structured exposure assessment, guideline-concordant molecular testing, human-supervised AI, and prospective exposure-resolved multi-omics studies.