Why Former Smokers Stay at High Lung Cancer Risk Decades After Quitting
New genomic review maps which molecular scars from smoking never fully heal—and how AI could turn them into prevention targets.
Résumé
Tobacco smoking leaves two categories of molecular damage in lung and airway tissue: changes that reverse within months of quitting, and changes that persist for decades. This comprehensive review synthesizes genomic, epigenomic, and transcriptomic evidence to distinguish these categories. Persistent alterations include mutations in TP53 and KRAS, aberrant DNA methylation at tumor suppressor loci, dysregulated noncoding RNAs, chromosomal instability, and epigenetic age acceleration. Reversible changes—like xenobiotic metabolism gene expression and acute inflammatory markers—normalize within months to two years. The authors argue that integrating multi-omics data with AI could produce composite molecular signatures to stratify high-risk former smokers, guide surveillance timing, and identify chemoprevention targets before irreversible damage drives malignancy.
Résumé détaillé
Lung cancer kills approximately 1.8 million people annually, and tobacco smoking accounts for roughly 85–90% of those deaths. Yet a striking epidemiological paradox persists: more than 40% of lung cancers in former smokers develop more than 15 years after cessation, and excess cancer risk remains over three times higher than in never-smokers even 25 years after quitting. Understanding why requires mapping which smoking-induced molecular alterations heal and which do not.
This review, drawing on longitudinal and cross-sectional studies of bronchial, nasal, and small airway epithelium, divides smoking-induced molecular changes into two functional categories. Nonpersistent (reversible) alterations include upregulation of xenobiotic metabolism genes such as CYP1A1, CYP1B1, and ALDH3A1, along with acute inflammatory cytokines and stress response markers. These normalize within weeks to roughly two years after cessation. Short-term cessation studies (3–6 months) already detect global shifts in DNA methylation at nearly 4,000 CpG sites, correlated with improved lung function, confirming that molecular recovery begins quickly.
Persistent alterations are more consequential for long-term cancer risk. Somatic mutations in driver genes TP53 and KRAS accumulate irreversibly, as do chromosomal copy number alterations and structural rearrangements that reflect decades of oxidative and carcinogen-induced DNA damage. Aberrant DNA methylation at tumor suppressor promoters—including CDKN2A/p16—can remain silenced for decades after cessation, with persistence appearing site-specific rather than simply proportional to cumulative smoking dose. Dysregulated microRNAs and long noncoding RNAs (lncRNAs) also show durable alterations in former smokers, contributing to sustained oncogenic signaling. Epigenetic clock analyses further reveal accelerated biological aging in former smokers that does not fully reverse, independent of chronological age.
Collectively, these persistent changes create a 'field cancerization' effect—widespread pre-neoplastic molecular reprogramming across the respiratory epithelium that remains years after the carcinogen source is removed. Polycyclic aromatic hydrocarbons drive G-to-T transversion mutations characteristic of squamous cell carcinoma, while tobacco-specific nitrosamines like NNK preferentially induce adenocarcinoma-associated alterations. Both patterns leave durable genomic imprints.
The authors propose that integrating multi-omics data (genomic, epigenomic, transcriptomic, proteomic, metabolomic) with AI and machine learning could transform this knowledge into clinical tools. Composite molecular signatures could stratify former smokers by residual cancer risk, optimize surveillance scheduling, and identify reversible epigenetic or transcriptomic targets for chemoprevention—potentially intercepting carcinogenesis before irreversible mutations accumulate. However, current challenges include small and heterogeneous cohorts, inconsistent temporal definitions of 'persistence,' limited longitudinal data with long follow-up, and poor generalizability of models across populations. The review calls for larger, prospectively designed multi-omics studies with standardized cessation timelines and AI validation across independent cohorts.
Principales conclusions
- Over 40% of lung cancers in former smokers arise more than 15 years post-cessation, confirming prolonged molecular risk.
- TP53 and KRAS somatic mutations, chromosomal instability, and CDKN2A methylation persist irreversibly after smoking cessation.
- Xenobiotic metabolism and acute inflammation gene expression normalize within weeks to ~2 years of quitting.
- Epigenetic age acceleration in former smokers does not fully reverse, representing a durable biological scar.
- AI integration of multi-omics data could enable precision risk stratification and chemoprevention targeting in former smokers.
Méthodologie
This is a narrative review synthesizing longitudinal and cross-sectional genomic, epigenomic, and transcriptomic studies of airway and lung tissues from smokers, former smokers, and never-smokers. Studies were selected through targeted biomedical database searches prioritizing persistence dynamics, tissue context, methodological rigor, and reproducibility across independent cohorts. No original data were generated; evidence was drawn from published cohort studies, epigenome-wide association studies, and AI-based cancer risk prediction research.
Limites de l'étude
Most underlying studies are small, heterogeneous, and cross-sectional rather than long-term longitudinal, limiting causal inference about persistence timelines. The definition of 'persistent' versus 'nonpersistent' varies across studies with no standardized cessation interval, making direct comparisons difficult. AI models developed in one population frequently fail to generalize to others, and current multi-omics integration approaches lack validation in large, prospective former-smoker cohorts.
Ce résumé vous a plu ?
Recevez les dernières recherches sur la longévité dans votre boîte de réception chaque semaine.
Saisissez votre e-mail pour vous abonner :