A new international NCI-led study offers new insights into the mutagenic forces leading to lung cancers in individuals who have never smoked. Published in Nature, the Sherlock-Lung study comprises the largest whole-genome sequencing (WGS) analysis to date of lung tumors in never-smokers and identifies both endogenous and exogenous mutational processes contributing to tumorigenesis.
“We know that up to 25% of all lung cancer cases globally occur in people who have never smoked,” said senior investigator Maria Teresa Landi, MD, PhD, NCI Integrative Tumor Epidemiology Branch (ITEB) and senior advisor for genomic epidemiology. “But in most of these patients, the known environmental risk factors—such as radon, asbestos, or secondhand smoke—don’t explain the disease. Our goal was to use mutational signatures as molecular fingerprints to infer cancer-causative processes and exposures.”
Study design and cohort composition
The Sherlock-Lung cohort includes WGS data from 871 treatment-naive lung cancer patients who self-identified as never-smokers, sampled from 28 geographically diverse regions. This global dataset allowed researchers to examine population-specific differences in tumor mutational landscapes and environmental exposures.
“Previous studies included only a few dozen never-smoker cases, often from single regions,” said Landi. “This is the first truly international and large-scale genomic dissection of lung cancer in this population.”
The cohort consisted primarily of lung adenocarcinomas, the most prevalent histological subtype in never-smokers. Researchers performed mutational signature analysis, telomere length estimation, and driver gene identification using high-depth whole-genome sequencing. Regional air pollution exposure was estimated based on PM2.5 (fine particulate matter) concentrations at the country and province/state level.
Key findings: endogenous processes and air pollution
One of the most significant findings was the dominance of a previously observed but still unexplained single base substitution signature, SBS40a, across much of the cohort. “This signature is likely driven by an endogenous process,” said Landi. “It accounted for the largest proportion of base substitutions and was enriched in tumors with EGFR mutations and in never-smokers relative to smokers. Its cause remains unknown, but it may represent a widespread, unrecognized mutagenic mechanism in multiple tissues.”
Importantly, the study found a robust association between air pollution and increased mutagenic burden. Patients from regions with high PM2.5 exposure had significantly higher rates of TP53 mutations, shorter telomeres, and increased activity of SBS4 (previously attributed to tobacco smoking) and SBS5, a clock-like signature.
“We observed a 3.9-fold increase in SBS4 and a 76% increase in SBS5 activity in tumors from patients in high-pollution areas,” Landi noted. “This strongly supports the idea that particulate air pollution acts both as a mutagen and a promoter of tumor development.”
In addition to elevated mutational burdens, high-pollution exposure correlated with telomere shortening, indicating possible proliferative stress and genomic instability. “Shortened telomeres reflect an increased number of cell divisions,” Landi explained. “In the context of air pollution, this suggests a dual mechanism—inducing DNA damage and promoting expansion of damaged clones.”
Aristolochic acid and region-specific mutagens
The study also uncovered SBS22a, a mutational signature associated with aristolochic acid exposure, exclusively in tumors from Taiwanese patients. Aristolochic acid, found in certain traditional herbal medicines, has been linked to bladder and kidney cancers, but this is the first evidence of its role in lung carcinogenesis.
“This was a surprising and striking finding,” said Landi. “The signature was almost completely restricted to Taiwan, highlighting the importance of regional environmental exposures and ethnopharmacological practices.”
Passive smoking exhibits weak mutagenicity
Among 458 patients with available secondhand smoke exposure data, researchers found minimal evidence of increased mutational burden or specific mutational signatures. While telomere shortening was observed in this group, suggesting enhanced cellular proliferation, secondhand smoke exposure did not correlate with the acquisition of canonical driver mutations or smoking-related mutational patterns.
“Secondhand smoke appears to have a much lower mutagenic potential than active smoking or air pollution,” Landi explained. “But it may still promote carcinogenesis by increasing cell turnover and clonal expansion.”
Evolutionary trajectories and clinical implications
The authors emphasize that lung cancer in never-smokers is not a monolithic entity but rather a collection of heterogeneous subtypes with distinct etiologies, evolutionary pathways, and potentially different responses to therapy.
“There’s considerable heterogeneity in the tumor evolutionary trajectories we’re beginning to characterize,” said Landi. “This has important implications for prognosis, early detection strategies, and therapeutic targeting. Ultimately, understanding these distinct molecular subtypes will allow us to move toward more precise interventions.”
Looking ahead
The Sherlock-Lung team is expanding their analysis to include additional regions with high environmental risk and will release the genomic dataset concurrent with the paper’s publication.
“We’re very open to collaboration,” said Landi. “This dataset is a resource for the entire community, and there’s much more to learn—especially about the unknown endogenous processes we’ve uncovered.”
The study was conducted in collaboration with Ludmil Alexandrov, PhD, and Marcos Díaz-Gay, PhD, at the University of California San Diego.
Landi concluded with a call to reframe how the scientific and clinical communities view lung cancer risk: “This work reinforces that lung cancer in never-smokers is not a residual category of smoking-related disease. It is a distinct and increasingly common cancer subtype shaped by diverse environmental and biological processes. Air pollution emerges here as a major, underrecognized carcinogen.”
