Lung Cancer Screening Must Move Beyond Smoking History, Review Argues

More than half of the world's lung cancer cases may never trigger a screening referral because eligibility is still built almost entirely around pack-year smoking history. A comprehensive review published in the American Society of Clinical Oncology Educational Book argued for scrapping that model in favor of a biology-driven approach that weighs air pollution, radon, diet, and genetics alongside tobacco exposure.¹

Screening Criteria Built for a Shrinking Population

Lung cancer remains the leading cause of cancer-related death worldwide, but the population it strikes has changed faster than screening guidelines have. Lung cancer in never-smokers (LCINS) accounts for an estimated 15% to 25% of cases globally, and in East Asian populations, particularly among women, it exceeds 60% of diagnoses.

In Taiwan, lung cancer has remained the leading cause of cancer death for more than 21 consecutive years despite a sustained, multidecade decline in smoking rates, a pattern the authors said reflects environmental and genetic drivers that pack-year thresholds cannot capture. Current eligibility criteria, including those from the US Preventive Services Task Force, may overlook more than half of the global lung cancer burden, disproportionately excluding nonsmoking women and East Asian populations.

The biology behind LCINS is still being filled in. A National Cancer Institute–led genomic analysis of tumors from never-smokers found that a majority carry mutational signatures tied to natural, endogenous processes rather than known external exposures, identifying 3 molecular subtypes with distinct growth patterns.² The exposome framework the present review proposes builds on that uncertainty, treating cumulative lifetime exposure to air, food, and water contaminants as a driver of carcinogenesis in its own right.¹

What the Exposome Adds to the Risk Picture

Fine particulate matter is an International Agency for Research on Cancer group 1 carcinogen that behaves less like a mutagen and more like a tumor promoter, triggering chronic lung inflammation that can push preexisting, age-related epidermal growth factor receptor mutations toward malignancy. Radon, an odorless radioactive gas, is the leading environmental cause of lung cancer among nonsmokers worldwide, with risk rising roughly 16% for every 100 Bq/m³ increase in long-term residential concentration and no identified safe threshold.

High-temperature cooking oil fumes, which are common in East Asian kitchens, as well as occupational exposures such as asbestos, silica, and diesel exhaust (roughly 15% of cases overall), and arsenic-contaminated groundwater round out the exposures the review said current screening criteria ignore entirely.

TALENT Cohort Shows What Broader Screening Could Catch

The Taiwanese Lung Cancer Screening in Never-Smokers (TALENT) study supplied the clinical evidence behind the review's case. The prospective cohort enrolled 12,011 never-smokers, half with a first-degree family history of lung cancer, and screened them with low-dose CT (LDCT).

The detection rate was 2.6%, well above the 1.1% and 0.9% reported in the smoking-history-based National Lung Screening Trial and NELSON trials, respectively. A total of 96.5% of cancers were caught at stage 0 or I, vs 50% to 60% diagnosed at stage IV in the general population. A first-degree family history of lung cancer was the strongest predictor in the cohort, with an OR of 4.5, ahead of any single environmental exposure measured.

Artificial Intelligence (AI) Risk Models and Molecular Triage

The review proposed pairing LDCT with a "dual-key" workflow: an AI-derived imaging score plus a molecular confirmation step. The deep-learning model Sybil, trained on a single 3-dimensional LDCT volume, reached an area under the receiver operating characteristic curve of 0.92 for 1-year risk prediction across multiple validation cohorts and can flag risk up to 6 years before diagnosis.

Circulating tumor DNA and methylation- and fragmentomic-based liquid biopsy assays were proposed as a high-confidence filter to spare patients biopsies for what often prove to be benign nodules. The same LDCT scan, the authors noted, already captures data on coronary artery calcium, emphysema, and vertebral bone density; the review argued for formalizing that into a comprehensive CT screening model that flags cardiovascular, respiratory, and bone health risk in one pass.

Uptake Gaps Persist Even Where Screening Exists

Better risk models will not close the gap alone. Only 18% to 22% of eligible US adults undergo LDCT screening, and disparities in insurance coverage, geographic access to imaging centers, and health literacy continue to delay diagnosis even where screening programs exist. The review frames this as a "humanomics" problem, in which structural inequities shape biological risk through chronic stress and inflammatory dysregulation, not just delayed access to a scanner.

"The unfair exclusion of the nonsmoking population cannot be any longer justified scientifically, ethically, or economically, with women in East Asia as the foremost example, alongside individuals living in high-pollution regions," the authors wrote.

For clinicians, payers, and health policy organizations, the argument is that risk-based eligibility, not smoking history alone, should determine who gets screened, with the TALENT data offered as evidence that broadening criteria finds cancer earlier, when lung-sparing surgery remains an option.

References

1. Gorría T, Marino N, Montoya Rubiano MA, Salinas KE, Flores EJ, Mezquita L, Yang PC. Lung cancer screening and prevention: beyond conventional strategies in a rapidly changing global milieu. Am Soc Clin Oncol Educ Book. 2026;46(3):e517088. doi:10.1200/EDBK-26-517088

2. NIH study illuminates origins of lung cancer in never smokers. National Institutes of Health. September 6, 2021. Accessed August 5, 2026. https://www.nih.gov/news-events/news-releases/nih-study-illuminates-origins-lung-cancer-never-smokers