Today, I review, link to, and excerpt from Trends In Cancer’s “Lung cancer in never smokers: from early detection to prevention”. [PubMed Abstract] [Full-Text HTML] [Full-Text PDF]. Trends Cancer. 2026 Apr;12(4):310-319. doi: 10.1016/j.trecan.2025.12.009. Epub 2026 Feb 11.
Bottom line: With a very few exceptions, there is no effective screening strategy for Lung cancer in never smokers.
Lung cancer in never smokers (LCINS) now accounts for a growing proportion of lung cancer cases, with distinct demographics and biology.
Identifying risk factors for LCINS remains challenging. The most well-established factors—germline variants, clonal hematopoiesis, and environmental exposures—currently lack validated screening tests.
Abstract
Lung cancer in never smokers (LCINS) is a growing global health challenge. Unlike smoking-related lung cancer, LCINS is characterized by distinct epidemiological patterns and unique molecular pathogenesis and, consequently, requires different clinical management approaches. Unfortunately, for patients with lung cancer who have never smoked, symptoms are nonspecific and often dismissed due to these patients not fitting a high-risk profile (e.g., smoker), underscoring the need for improved detection and interception. Emerging risk factors, including germline variants, clonal hematopoiesis, and environmental exposures, offer new avenues for risk stratification and preventive strategies. While low-dose computed tomography screening shows promise in high-risk subgroups, challenges remain in optimizing cost-effectiveness. Novel prevention approaches, from interleukin (IL)-1β inhibition to cancer vaccines, are under investigation. This opinion article discusses why LCINS demands unique clinical and research paradigms to address its biological complexity.
Lung cancer is the leading cause of cancer-related deaths worldwide [1]. As smoking rates have decreased, lung cancer in never smokers (LCINS) (see Glossary) [2,3] accounts for an increasing proportion of patients [1]. Recent findings suggest that the absolute incidence of LCINS may be increasing [4,5]. Unfortunately, a higher proportion of LCINS cases goes undiagnosed for longer periods compared to those involving smoking-related lung cancer, and many patients are diagnosed with metastatic disease and therefore receive treatment with palliative intent [3,6]. In part, this relates to the lack of lung cancer screening programs for never smokers, as the background risk is low (incidence 5–20/100 000) [7], and there are no established risk algorithms other than age to stratify individuals for screening. LCINS often harbors mutations or fusions in epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK). Results from trials of adjuvant tyrosine kinase inhibitor (TKI) treatment following surgery in patients with EGFR and ALK mutation-positive lung cancer show a benefit even in the earliest stages of disease (stage IB and II), prior to detection of metastatic dissemination [8,9]. The recurrence patterns observed in patients across these studies indicate that adding adjuvant TKIs protects against extrathoracic and central nervous system metastases. These findings underscore the urgent need for earlier detection and interception strategies to increase the proportion of patients whose cancer is detected at the earliest possible stage and ultimately to reduce lung cancer mortality.
From a clinical point of view, never smokers are more likely to present with adenocarcinoma histology and oncogene-addicted tumors (especially for EGFR and ALK), making them candidates for targeted therapies [11]. Multiple studies have demonstrated that a high prevalence of actionable driver mutations (~80%) is found in LCINS, in contrast to LCS [1,12–15]. Pivotal immune checkpoint inhibitor (ICI) trials consistently show that patients with lung cancer who have never smoked, unlike those who have smoked, are significantly less responsive to ICI regimens, partly because LCINS tumors have a lower mutational burden [16]. These findings reinforce the concept that LCINS represents a distinct biological entity compared to patients with a smoking history.
Lung cancer prevention is an area still in its infancy. Studies testing therapies at the earliest stages of disease, even prior to transformation, are being carried out in both preclinical and clinical settings. These include targeted therapies, immunotherapies, and cancer vaccines. All these approaches could have major benefits for patients, but these benefits must outweigh the risk of toxicity, iatrogenic complications, and economic costs to health care systems. Targeted therapies, such as EGFR and ALK TKIs and KRAS G12C inhibitors, commonly cause dermatologic and gastrointestinal toxicities. These side effects are usually low grade but can significantly impact patient quality of life [17–19]. Immunotherapies can cause nausea, fatigue, and diarrhea, as well as immune-related adverse events such as pneumonitis, which can be life-threatening [20–23]. Cancer vaccines, in general, have a higher safety profile than both targeted therapies and immunotherapies, with some studies showing only mild to moderate adverse events [24,25]. For all these therapies, evidence of benefits that outweigh risks is needed in the relevant patient population—individuals at high risk of lung cancer or with preneoplastic lung lesions [24–26].
Risk factors for LCINS
Identifying those at risk of lung cancer in a never-smoker population is particularly challenging. Environmental and genetic factors have been implicated in the etiology of lung cancer in nonsmoking populations. These include germline variants in oncogenes and DNA damage response genes (e.g., EGFR, TP53, and ATM, among others) [27–29], clonal hematopoiesis [30], radon exposure [31], secondhand smoke [1,32–35], inflammatory diseases [36], and air pollution [37] (Figure 1). However, much of the supporting evidence is retrospective, and the relative risk associated with each of these exposures is modest, so justifying costly screening or therapy is difficult. Together, the global disease burden from each of these factors results in a significant number of deaths annually [38] (https://www.healthdata.org/research-analysis/library/transport-health-global-burden-disease-motorized-road-transport), supporting more work aimed at lung cancer prevention in a never-smoker population.
Figure 1Schematic of the innate and extrinsic risk factors for lung cancer in patients who have never smoked.
Inherited or acquired predisposition
Germline alterations
It is estimated that up to 4.5% of patients with lung adenocarcinoma have germline variants that correlate with increased lung cancer risk [39]. Specific germline mutations in known oncogenes, such as EGFR p.Thr790Met (T790M), have been associated with increased lung cancer risk (>50% diagnosed with lung cancer by age 60) [40,41]. Individuals with these mutations are at high risk of developing multifocal lung lesions consistent with adenocarcinoma spectrum diagnosis, even at a very early age, with a high risk of developing invasive EGFR-driven lung cancer [40,41]. In addition to canonical germline variants, early developmental mosaicism has recently been investigated as a distinct mechanism of genetic predisposition to EGFR-mutant lung cancer [42].
Germline alterations in the apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3 (APOBEC3) family of cytosine deaminases are also linked with elevated lung cancer risk [27–29]. APOBEC3 genes normally function as part of the body’s defense against viruses, attacking and mutating viral RNA and DNA [43]. Unfortunately, APOBEC3 genes can also mutate the host genome and are linked with DNA damage and mutational signatures in up to 70% of human cancer types [43]. The APOBEC3A/B germline deletion correlates with significantly increased non-small cell lung cancer (NSCLC) risk, with a 2.7-fold increase in risk in a southern Chinese population after adjusting for age, sex, and smoking history [29]. In another study based on a Norwegian population, the APOBEC3A/B germline deletion polymorphism was associated with a younger age at diagnosis of both lung and prostate cancer, with a 2.2-fold increase in risk [27]. The impact of this germline alteration is likely population-dependent. In an East Asian population (APOBEC3A/3B prevalence 37%), about 15% of NSCLC cases were attributable to the APOBEC3A/B germline deletion, while in Native American populations, where the APOBEC3A/B germline deletion is more prevalent (57.7%), the impact was greater, with about 30% of NSCLC cases potentially attributable to the APOBEC3A/B germline deletion [44]. Screening for the APOBEC3A/B germline deletion polymorphism is simple and low cost, as it uses polymerase chain reaction [27,44]. [Emphasis added]
Clonal hematopoiesis
Another potential at-risk population comprises individuals with an expansion of somatic alterations in the DNA of hematopoietic stem cells, called clonal hematopoiesis of indeterminate potential (CHIP). CHIP is associated with an elevated risk of developing several types of solid cancers [45,46]. Recent data demonstrate that CHIP carriers with a high variant allele fraction (≥ 0.1) have an increased risk of lung cancer, independent of smoking status [30,45,46]. CHIP alterations significantly associated with increased lung cancer include the most frequently mutated CHIP genes—DNMT3A, TET2, and ASXL1[30,45,46]. The mechanism behind the increased risk of lung cancer with CHIP is complex. Research points to the involvement of increased inflammatory processes [47,48]. Macrophages show elevated secretion of cytokines and chemokines, including CXCL1, CXCL2, CXCL3, CXCL4, IL-1β, and IL-16 [47–49]. In mouse models, inhibiting IL-1β reduces lung cancer [37], while decreased lung cancer incidence was observed in patients treated with a therapeutic monoclonal antibody targeting IL-1β [50,51]. However, recent evidence suggests that the effect may not be as significant as initially observed [52]. Despite these conflicting findings, anti-IL-1β therapy could be an approach for lung cancer prevention in individuals classified as high risk. Studies have also revealed that tumor-infiltrating clonal hematopoiesis increases the risk of lung cancer progression by driving tumor immune evasion and growth, underscoring how age-related clonal hematopoiesis fuels cancer evolution [45]. However, there are currently no standard guidelines for screening cancer patients or those at high risk of cancer for CHIP [53].
Highlights
Lung cancer in never smokers (LCINS) now accounts for a growing proportion of lung cancer cases, with distinct demographics and biology.
Identifying risk factors for LCINS remains challenging. The most well-established factors—germline variants, clonal hematopoiesis, and environmental exposures—currently lack validated screening tests.
Abstract
Lung cancer in never smokers (LCINS) is a growing global health challenge. Unlike smoking-related lung cancer, LCINS is characterized by distinct epidemiological patterns and unique molecular pathogenesis and, consequently, requires different clinical management approaches. Unfortunately, for patients with lung cancer who have never smoked, symptoms are nonspecific and often dismissed due to these patients not fitting a high-risk profile (e.g., smoker), underscoring the need for improved detection and interception. Emerging risk factors, including germline variants, clonal hematopoiesis, and environmental exposures, offer new avenues for risk stratification and preventive strategies. While low-dose computed tomography screening shows promise in high-risk subgroups, challenges remain in optimizing cost-effectiveness. Novel prevention approaches, from interleukin (IL)-1β inhibition to cancer vaccines, are under investigation. This opinion article discusses why LCINS demands unique clinical and research paradigms to address its biological complexity.
Lung cancer is the leading cause of cancer-related deaths worldwide [1]. As smoking rates have decreased, lung cancer in never smokers (LCINS) (see Glossary) [2,3] accounts for an increasing proportion of patients [1]. Recent findings suggest that the absolute incidence of LCINS may be increasing [4,5]. Unfortunately, a higher proportion of LCINS cases goes undiagnosed for longer periods compared to those involving smoking-related lung cancer, and many patients are diagnosed with metastatic disease and therefore receive treatment with palliative intent [3,6]. In part, this relates to the lack of lung cancer screening programs for never smokers, as the background risk is low (incidence 5–20/100 000) [7], and there are no established risk algorithms other than age to stratify individuals for screening. LCINS often harbors mutations or fusions in epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK). Results from trials of adjuvant tyrosine kinase inhibitor (TKI) treatment following surgery in patients with EGFR and ALK mutation-positive lung cancer show a benefit even in the earliest stages of disease (stage IB and II), prior to detection of metastatic dissemination [8,9]. The recurrence patterns observed in patients across these studies indicate that adding adjuvant TKIs protects against extrathoracic and central nervous system metastases. These findings underscore the urgent need for earlier detection and interception strategies to increase the proportion of patients whose cancer is detected at the earliest possible stage and ultimately to reduce lung cancer mortality.
The demographics of patients with lung cancer who have never smoked differ markedly from those of lung cancer in patients who have smoked (LCS). LCINS occurs more commonly in women and Asian populations, with women who have never smoked being more than twice as likely to get lung cancer as men who have never smoked [1,9,10]. However, this increased risk in women remains controversial and poorly understood [1,3]. A better understanding of the biological mechanisms driving LCINS is essential not only to inform cancer prevention strategies and improve patient outcomes but also to enable the identification of high-risk individuals who would benefit most from targeted screening approaches.
From a clinical point of view, never smokers are more likely to present with adenocarcinoma histology and oncogene-addicted tumors (especially for EGFR and ALK), making them candidates for targeted therapies [11]. Multiple studies have demonstrated that a high prevalence of actionable driver mutations (~80%) is found in LCINS, in contrast to LCS [1,12–15]. Pivotal immune checkpoint inhibitor (ICI) trials consistently show that patients with lung cancer who have never smoked, unlike those who have smoked, are significantly less responsive to ICI regimens, partly because LCINS tumors have a lower mutational burden [16]. These findings reinforce the concept that LCINS represents a distinct biological entity compared to patients with a smoking history.
Lung cancer prevention is an area still in its infancy. Studies testing therapies at the earliest stages of disease, even prior to transformation, are being carried out in both preclinical and clinical settings. These include targeted therapies, immunotherapies, and cancer vaccines. All these approaches could have major benefits for patients, but these benefits must outweigh the risk of toxicity, iatrogenic complications, and economic costs to health care systems. Targeted therapies, such as EGFR and ALK TKIs and KRAS G12C inhibitors, commonly cause dermatologic and gastrointestinal toxicities. These side effects are usually low grade but can significantly impact patient quality of life [17–19]. Immunotherapies can cause nausea, fatigue, and diarrhea, as well as immune-related adverse events such as pneumonitis, which can be life-threatening [20–23]. Cancer vaccines, in general, have a higher safety profile than both targeted therapies and immunotherapies, with some studies showing only mild to moderate adverse events [24,25]. For all these therapies, evidence of benefits that outweigh risks is needed in the relevant patient population—individuals at high risk of lung cancer or with preneoplastic lung lesions [24–26].
Risk factors for LCINS
Identifying those at risk of lung cancer in a never-smoker population is particularly challenging. Environmental and genetic factors have been implicated in the etiology of lung cancer in nonsmoking populations. These include germline variants in oncogenes and DNA damage response genes (e.g., EGFR, TP53, and ATM, among others) [27–29], clonal hematopoiesis [30], radon exposure [31], secondhand smoke [1,32–35], inflammatory diseases [36], and air pollution [37] (Figure 1). However, much of the supporting evidence is retrospective, and the relative risk associated with each of these exposures is modest, so justifying costly screening or therapy is difficult. Together, the global disease burden from each of these factors results in a significant number of deaths annually [38] (https://www.healthdata.org/research-analysis/library/transport-health-global-burden-disease-motorized-road-transport), supporting more work aimed at lung cancer prevention in a never-smoker population.
Figure 1Schematic of the innate and extrinsic risk factors for lung cancer in patients who have never smoked.
Inherited or acquired predisposition
Germline alterations
It is estimated that up to 4.5% of patients with lung adenocarcinoma have germline variants that correlate with increased lung cancer risk [39]. Specific germline mutations in known oncogenes, such as EGFR p.Thr790Met (T790M), have been associated with increased lung cancer risk (>50% diagnosed with lung cancer by age 60) [40,41]. Individuals with these mutations are at high risk of developing multifocal lung lesions consistent with adenocarcinoma spectrum diagnosis, even at a very early age, with a high risk of developing invasive EGFR-driven lung cancer [40,41]. In addition to canonical germline variants, early developmental mosaicism has recently been investigated as a distinct mechanism of genetic predisposition to EGFR-mutant lung cancer [42].
Germline alterations in the apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3 (APOBEC3) family of cytosine deaminases are also linked with elevated lung cancer risk [27–29]. APOBEC3 genes normally function as part of the body’s defense against viruses, attacking and mutating viral RNA and DNA [43]. Unfortunately, APOBEC3 genes can also mutate the host genome and are linked with DNA damage and mutational signatures in up to 70% of human cancer types [43]. The APOBEC3A/B germline deletion correlates with significantly increased non-small cell lung cancer (NSCLC) risk, with a 2.7-fold increase in risk in a southern Chinese population after adjusting for age, sex, and smoking history [29]. In another study based on a Norwegian population, the APOBEC3A/B germline deletion polymorphism was associated with a younger age at diagnosis of both lung and prostate cancer, with a 2.2-fold increase in risk [27]. The impact of this germline alteration is likely population-dependent. In an East Asian population (APOBEC3A/3B prevalence 37%), about 15% of NSCLC cases were attributable to the APOBEC3A/B germline deletion, while in Native American populations, where the APOBEC3A/B germline deletion is more prevalent (57.7%), the impact was greater, with about 30% of NSCLC cases potentially attributable to the APOBEC3A/B germline deletion [44]. Screening for the APOBEC3A/B germline deletion polymorphism is simple and low cost, as it uses polymerase chain reaction [27,44]. [Emphasis Added]
Clonal hematopoiesis
Another potential at-risk population comprises individuals with an expansion of somatic alterations in the DNA of hematopoietic stem cells, called clonal hematopoiesis of indeterminate potential (CHIP). CHIP is associated with an elevated risk of developing several types of solid cancers [45,46]. Recent data demonstrate that CHIP carriers with a high variant allele fraction (≥ 0.1) have an increased risk of lung cancer, independent of smoking status [30,45,46]. CHIP alterations significantly associated with increased lung cancer include the most frequently mutated CHIP genes—DNMT3A, TET2, and ASXL1[30,45,46]. The mechanism behind the increased risk of lung cancer with CHIP is complex. Research points to the involvement of increased inflammatory processes [47,48]. Macrophages show elevated secretion of cytokines and chemokines, including CXCL1, CXCL2, CXCL3, CXCL4, IL-1β, and IL-16 [47–49]. In mouse models, inhibiting IL-1β reduces lung cancer [37], while decreased lung cancer incidence was observed in patients treated with a therapeutic monoclonal antibody targeting IL-1β [50,51]. However, recent evidence suggests that the effect may not be as significant as initially observed [52]. Despite these conflicting findings, anti-IL-1β therapy could be an approach for lung cancer prevention in individuals classified as high risk. Studies have also revealed that tumor-infiltrating clonal hematopoiesis increases the risk of lung cancer progression by driving tumor immune evasion and growth, underscoring how age-related clonal hematopoiesis fuels cancer evolution [45]. However, there are currently no standard guidelines for screening cancer patients or those at high risk of cancer for CHIP [53].
The exposome
Radiation exposure
Exposure to radon, a radioactive gas, is a risk factor for lung cancer development. As a decay product of uranium-238, radon can accumulate in poorly ventilated buildings. Because radon levels fluctuate, exposure is difficult to assess; however, geographic mapping has identified areas with persistently elevated levels, enabling targeted public health interventions and risk mitigation [31]. Initial studies of radon and lung cancer focused on miners with high exposures and found that cumulative radon exposure was associated with increased lung cancer risk [31,54]. Researchers have investigated links between radon exposure and specific genomic alterations that drive lung cancer, but findings are mixed [31]. In vitro studies have linked radiation exposure to mutations in the tumor suppressor TP53, and in rats, radon exposure has induced chromosome losses in regions harboring tumor suppressor genes [31]. More work is required to understand the mechanism behind the increased risk of lung cancer from radon exposure. Without this knowledge, prevention through monitoring is the most effective strategy [31,54,55].
In developed countries, the expanded use of CT scanners has substantially increased diagnostic exposure to ionizing radiation. Recent epidemiological modeling studies suggest that a large proportion of malignancies could be a result of diagnostic imaging exposure [56]. A recent examination of mutational signatures in the cancer genomes from patients with lung cancer who have never smoked highlighted that 87% of indel drivers in EGFR (predominantly exon 19 deletions) occur in the context of the mutational signature related to radiation exposure [57]. A cumulative estimate of a person’s exposure to environmental and medical radiation could inform risk stratification for screening or interception strategies.
Second-hand smoke
Exposure to second-hand smoke has been identified as a significant risk factor (estimated to increase risk by 20–25%) for the development of LCINS [1,32–35]. Although patients with lung cancer who have never smoked who were exposed to second-hand smoke have driver mutation profiles (EGFR, ALK, Kirsten rat sarcoma viral oncogene homolog (KRAS), Human epidermal growth factor receptor 2 (HER2), B-Raf proto-oncogene, serine/threonine kinase (BRAF), and Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA)) similar to those unexposed, second-hand smoke exposure still increases lung cancer risk by 20–25% [1]. Genomic studies of tumors in LCINS with second-hand smoke exposure also do not show significant increases in smoking-related mutational signatures [1,57]. One recent study showed a small but statistically significant increase in tumor mutational burden (TMB) in lung cancer patients exposed to second-hand smoke, but this increase in mutational burden was not due to classic smoking signatures (SBS4a and SBS4b). Instead, it was associated with subclonal increases in APOBEC-related signatures (SBS2 and SBS13) [35]. A recent preprint also demonstrated a synergistic effect between tobacco exposure and APOBEC-related mutation signature accumulation in mouse models [35,58]. These findings together illustrate the importance of second-hand smoke exposure as a risk factor for never-smoker lung cancer, but potentially not as the sole cause of LCINS.
Air pollution
Particulate matter air pollution with a diameter less than 2.5 μm (PM2.5) is classified as a carcinogen and a cause of lung cancer by the International Agency for Research on Cancer [59]. Recent analysis demonstrates a strong association between PM2.5 levels and lung cancer incidence [31,60,61]. Emerging data demonstrate increased mutation burden, including single-base substitution, double-base substitution, and indel mutations, as well as telomere shortening and increased incidence of TP53 mutations in patients diagnosed with lung cancer in areas with high PM2.5 levels [57]. Using genetically engineered mouse models, air pollution was found to drive an increase in macrophages and IL-1β release in the lung, fueling tumorigenesis [61], similar to the inflammatory process observed in patients with CHIP [48,49].
Identifying individuals at risk of lung cancer due to air pollution is challenging but could be achieved by classifying individuals by postcode based on air pollution levels. An alternative approach could involve targeting the inflammatory processes linked with air pollution in those individuals with high exposure. This approach is beginning to be explored in preclinical models of lung cancer [62]. However, lifetime exposure to air pollution is the sum of both indoor and outdoor exposures. Indoor air pollution (e.g., open fires, solid fuel stoves, poorly ventilated kitchens) is understudied, which makes risk stratification outside of urban environments, particularly in South Asia, parts of East and Southeast Asia, Africa, and some urban centers in the Middle East and South America, challenging.
Inflammatory conditions
Recent work has suggested an association between certain inflammatory conditions, such as gastroesophageal reflux disease, upper respiratory infections, type 1 diabetes mellitus, chronic gastrointestinal conditions, and viral infections, with an increased risk of lung cancer. However, the observational nature of these studies and potential confounding factors limit definitive conclusions about causality [36,63]. As for pulmonary inflammatory diseases, emerging evidence indicates that chronic obstructive pulmonary disease may represent a smoking-independent risk factor for lung cancer, whereas the association appears less definitive for alpha-1 antitrypsin deficiency [64,65].
Screening for early detection of LCINS
Detection of lung cancer at its earliest stages has a significant impact on patient mortality [66]. The National Lung Cancer Screening Trial (NLST), which used low-dose computed tomography (LDCT) to screen for lung cancer in high-risk individuals, demonstrated a 20% reduction in lung cancer-specific mortality [66,67]. High-risk individuals were defined as those aged 55–74 years with a smoking history of at least 30 pack-years [67]. Additional clinical trials of patients screened with LDCT have shown decreased lung cancer-related mortality, increased detection of early-stage lung cancer, and reduced late-stage disease [10,67,68]. However, without first selecting a high-risk population, these programs can cause unnecessary harm to participants, with false positives resulting in patient anxiety as well as unnecessary invasive procedures [66]. From an economic perspective, recent analyses suggest that LDCT screening can be cost-effective when targeted to high-risk groups [69]. Defining high-risk individuals in a never-smoker population is more difficult than in a smoker population, and the evidence of benefit from screening in never smokers is controversial [70,71].
Recent efforts to expand lung cancer screening to light- and never-smoker populations have proven complicated, as it is unclear if screening results in improved overall survival [71]. The recent Taiwan Lung Cancer Screening in Never-Smoker Trial (TALENT) highlighted the increased risk of lung cancer in never smokers with a family history of lung cancer in a first-degree relative [70]. As a result, the Taiwan National Screening Program now includes patients without a smoking history when there is a family history of lung cancer, with a 1.2 and 1.8% detection rate in males and females, respectively [70]. One systematic review and meta-analysis also demonstrated that lung cancer detection in female never smokers was just as effective as detection in male ever smokers in Asia [2]. These data together suggest that when a high-risk population is screened, in this case Asian female never smokers with a family history of lung cancer, LDCT can detect early-stage lung cancer and result in significant reductions in lung cancer-specific mortality. In an ongoing trial, Asian female never smokers based in New York (New York Female Asian Nonsmoker Screening Study) are being screened with LDCT. Preliminary data from this trial suggest that LDCT screening in this patient population is feasible, and the invasive adenocarcinoma detection rate is higher than in the NLST [72]. The data support further studies of whether populations with a family history of lung cancer could benefit from LDCT screening. Currently, there are no randomized trials to assess the survival and health economic benefit of LDCT in a never-smoker population.
Multi-cancer early detection testing is another approach for early lung cancer detection, with the aim to identify signals from multiple cancer types through cell-free DNA using a simple blood test [73–75]. So far, detection of early-stage disease has proven challenging [73,76]. These assays have shown great promise, even though tumor-informed panels seem to perform better in terms of tumor and recurrence detection [77,78].
Highlights
Lung cancer in never smokers (LCINS) now accounts for a growing proportion of lung cancer cases, with distinct demographics and biology.
Identifying risk factors for LCINS remains challenging. The most well-established factors—germline variants, clonal hematopoiesis, and environmental exposures—currently lack validated screening tests.
Abstract
Lung cancer in never smokers (LCINS) is a growing global health challenge. Unlike smoking-related lung cancer, LCINS is characterized by distinct epidemiological patterns and unique molecular pathogenesis and, consequently, requires different clinical management approaches. Unfortunately, for patients with lung cancer who have never smoked, symptoms are nonspecific and often dismissed due to these patients not fitting a high-risk profile (e.g., smoker), underscoring the need for improved detection and interception. Emerging risk factors, including germline variants, clonal hematopoiesis, and environmental exposures, offer new avenues for risk stratification and preventive strategies. While low-dose computed tomography screening shows promise in high-risk subgroups, challenges remain in optimizing cost-effectiveness. Novel prevention approaches, from interleukin (IL)-1β inhibition to cancer vaccines, are under investigation. This opinion article discusses why LCINS demands unique clinical and research paradigms to address its biological complexity.
Lung cancer is the leading cause of cancer-related deaths worldwide [1]. As smoking rates have decreased, lung cancer in never smokers (LCINS) (see Glossary) [2,3] accounts for an increasing proportion of patients [1]. Recent findings suggest that the absolute incidence of LCINS may be increasing [4,5]. Unfortunately, a higher proportion of LCINS cases goes undiagnosed for longer periods compared to those involving smoking-related lung cancer, and many patients are diagnosed with metastatic disease and therefore receive treatment with palliative intent [3,6]. In part, this relates to the lack of lung cancer screening programs for never smokers, as the background risk is low (incidence 5–20/100 000) [7], and there are no established risk algorithms other than age to stratify individuals for screening. LCINS often harbors mutations or fusions in epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK). Results from trials of adjuvant tyrosine kinase inhibitor (TKI) treatment following surgery in patients with EGFR and ALK mutation-positive lung cancer show a benefit even in the earliest stages of disease (stage IB and II), prior to detection of metastatic dissemination [8,9]. The recurrence patterns observed in patients across these studies indicate that adding adjuvant TKIs protects against extrathoracic and central nervous system metastases. These findings underscore the urgent need for earlier detection and interception strategies to increase the proportion of patients whose cancer is detected at the earliest possible stage and ultimately to reduce lung cancer mortality.
The demographics of patients with lung cancer who have never smoked differ markedly from those of lung cancer in patients who have smoked (LCS). LCINS occurs more commonly in women and Asian populations, with women who have never smoked being more than twice as likely to get lung cancer as men who have never smoked [1,9,10]. However, this increased risk in women remains controversial and poorly understood [1,3]. A better understanding of the biological mechanisms driving LCINS is essential not only to inform cancer prevention strategies and improve patient outcomes but also to enable the identification of high-risk individuals who would benefit most from targeted screening approaches.
From a clinical point of view, never smokers are more likely to present with adenocarcinoma histology and oncogene-addicted tumors (especially for EGFR and ALK), making them candidates for targeted therapies [11]. Multiple studies have demonstrated that a high prevalence of actionable driver mutations (~80%) is found in LCINS, in contrast to LCS [1,12–15]. Pivotal immune checkpoint inhibitor (ICI) trials consistently show that patients with lung cancer who have never smoked, unlike those who have smoked, are significantly less responsive to ICI regimens, partly because LCINS tumors have a lower mutational burden [16]. These findings reinforce the concept that LCINS represents a distinct biological entity compared to patients with a smoking history.
Lung cancer prevention is an area still in its infancy. Studies testing therapies at the earliest stages of disease, even prior to transformation, are being carried out in both preclinical and clinical settings. These include targeted therapies, immunotherapies, and cancer vaccines. All these approaches could have major benefits for patients, but these benefits must outweigh the risk of toxicity, iatrogenic complications, and economic costs to health care systems. Targeted therapies, such as EGFR and ALK TKIs and KRAS G12C inhibitors, commonly cause dermatologic and gastrointestinal toxicities. These side effects are usually low grade but can significantly impact patient quality of life [17–19]. Immunotherapies can cause nausea, fatigue, and diarrhea, as well as immune-related adverse events such as pneumonitis, which can be life-threatening [20–23]. Cancer vaccines, in general, have a higher safety profile than both targeted therapies and immunotherapies, with some studies showing only mild to moderate adverse events [24,25]. For all these therapies, evidence of benefits that outweigh risks is needed in the relevant patient population—individuals at high risk of lung cancer or with preneoplastic lung lesions [24–26].
Risk factors for LCINS
Identifying those at risk of lung cancer in a never-smoker population is particularly challenging. Environmental and genetic factors have been implicated in the etiology of lung cancer in nonsmoking populations. These include germline variants in oncogenes and DNA damage response genes (e.g., EGFR, TP53, and ATM, among others) [27–29], clonal hematopoiesis [30], radon exposure [31], secondhand smoke [1,32–35], inflammatory diseases [36], and air pollution [37] (Figure 1). However, much of the supporting evidence is retrospective, and the relative risk associated with each of these exposures is modest, so justifying costly screening or therapy is difficult. Together, the global disease burden from each of these factors results in a significant number of deaths annually [38] (https://www.healthdata.org/research-analysis/library/transport-health-global-burden-disease-motorized-road-transport), supporting more work aimed at lung cancer prevention in a never-smoker population.
Figure 1Schematic of the innate and extrinsic risk factors for lung cancer in patients who have never smoked.
Inherited or acquired predisposition
Germline alterations
It is estimated that up to 4.5% of patients with lung adenocarcinoma have germline variants that correlate with increased lung cancer risk [39]. Specific germline mutations in known oncogenes, such as EGFR p.Thr790Met (T790M), have been associated with increased lung cancer risk (>50% diagnosed with lung cancer by age 60) [40,41]. Individuals with these mutations are at high risk of developing multifocal lung lesions consistent with adenocarcinoma spectrum diagnosis, even at a very early age, with a high risk of developing invasive EGFR-driven lung cancer [40,41]. In addition to canonical germline variants, early developmental mosaicism has recently been investigated as a distinct mechanism of genetic predisposition to EGFR-mutant lung cancer [42].
Germline alterations in the apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3 (APOBEC3) family of cytosine deaminases are also linked with elevated lung cancer risk [27–29]. APOBEC3 genes normally function as part of the body’s defense against viruses, attacking and mutating viral RNA and DNA [43]. Unfortunately, APOBEC3 genes can also mutate the host genome and are linked with DNA damage and mutational signatures in up to 70% of human cancer types [43]. The APOBEC3A/B germline deletion correlates with significantly increased non-small cell lung cancer (NSCLC) risk, with a 2.7-fold increase in risk in a southern Chinese population after adjusting for age, sex, and smoking history [29]. In another study based on a Norwegian population, the APOBEC3A/B germline deletion polymorphism was associated with a younger age at diagnosis of both lung and prostate cancer, with a 2.2-fold increase in risk [27]. The impact of this germline alteration is likely population-dependent. In an East Asian population (APOBEC3A/3B prevalence 37%), about 15% of NSCLC cases were attributable to the APOBEC3A/B germline deletion, while in Native American populations, where the APOBEC3A/B germline deletion is more prevalent (57.7%), the impact was greater, with about 30% of NSCLC cases potentially attributable to the APOBEC3A/B germline deletion [44]. Screening for the APOBEC3A/B germline deletion polymorphism is simple and low cost, as it uses polymerase chain reaction [27,44].
Clonal hematopoiesis
Another potential at-risk population comprises individuals with an expansion of somatic alterations in the DNA of hematopoietic stem cells, called clonal hematopoiesis of indeterminate potential (CHIP). CHIP is associated with an elevated risk of developing several types of solid cancers [45,46]. Recent data demonstrate that CHIP carriers with a high variant allele fraction (≥ 0.1) have an increased risk of lung cancer, independent of smoking status [30,45,46]. CHIP alterations significantly associated with increased lung cancer include the most frequently mutated CHIP genes—DNMT3A, TET2, and ASXL1[30,45,46]. The mechanism behind the increased risk of lung cancer with CHIP is complex. Research points to the involvement of increased inflammatory processes [47,48]. Macrophages show elevated secretion of cytokines and chemokines, including CXCL1, CXCL2, CXCL3, CXCL4, IL-1β, and IL-16 [47–49]. In mouse models, inhibiting IL-1β reduces lung cancer [37], while decreased lung cancer incidence was observed in patients treated with a therapeutic monoclonal antibody targeting IL-1β [50,51]. However, recent evidence suggests that the effect may not be as significant as initially observed [52]. Despite these conflicting findings, anti-IL-1β therapy could be an approach for lung cancer prevention in individuals classified as high risk. Studies have also revealed that tumor-infiltrating clonal hematopoiesis increases the risk of lung cancer progression by driving tumor immune evasion and growth, underscoring how age-related clonal hematopoiesis fuels cancer evolution [45]. However, there are currently no standard guidelines for screening cancer patients or those at high risk of cancer for CHIP [53].
The exposome
Radiation exposure
Exposure to radon, a radioactive gas, is a risk factor for lung cancer development. As a decay product of uranium-238, radon can accumulate in poorly ventilated buildings. Because radon levels fluctuate, exposure is difficult to assess; however, geographic mapping has identified areas with persistently elevated levels, enabling targeted public health interventions and risk mitigation [31]. Initial studies of radon and lung cancer focused on miners with high exposures and found that cumulative radon exposure was associated with increased lung cancer risk [31,54]. Researchers have investigated links between radon exposure and specific genomic alterations that drive lung cancer, but findings are mixed [31]. In vitro studies have linked radiation exposure to mutations in the tumor suppressor TP53, and in rats, radon exposure has induced chromosome losses in regions harboring tumor suppressor genes [31]. More work is required to understand the mechanism behind the increased risk of lung cancer from radon exposure. Without this knowledge, prevention through monitoring is the most effective strategy [31,54,55].
In developed countries, the expanded use of CT scanners has substantially increased diagnostic exposure to ionizing radiation. Recent epidemiological modeling studies suggest that a large proportion of malignancies could be a result of diagnostic imaging exposure [56]. A recent examination of mutational signatures in the cancer genomes from patients with lung cancer who have never smoked highlighted that 87% of indel drivers in EGFR (predominantly exon 19 deletions) occur in the context of the mutational signature related to radiation exposure [57]. A cumulative estimate of a person’s exposure to environmental and medical radiation could inform risk stratification for screening or interception strategies.
Second-hand smoke
Exposure to second-hand smoke has been identified as a significant risk factor (estimated to increase risk by 20–25%) for the development of LCINS [1,32–35]. Although patients with lung cancer who have never smoked who were exposed to second-hand smoke have driver mutation profiles (EGFR, ALK, Kirsten rat sarcoma viral oncogene homolog (KRAS), Human epidermal growth factor receptor 2 (HER2), B-Raf proto-oncogene, serine/threonine kinase (BRAF), and Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA)) similar to those unexposed, second-hand smoke exposure still increases lung cancer risk by 20–25% [1]. Genomic studies of tumors in LCINS with second-hand smoke exposure also do not show significant increases in smoking-related mutational signatures [1,57]. One recent study showed a small but statistically significant increase in tumor mutational burden (TMB) in lung cancer patients exposed to second-hand smoke, but this increase in mutational burden was not due to classic smoking signatures (SBS4a and SBS4b). Instead, it was associated with subclonal increases in APOBEC-related signatures (SBS2 and SBS13) [35]. A recent preprint also demonstrated a synergistic effect between tobacco exposure and APOBEC-related mutation signature accumulation in mouse models [35,58]. These findings together illustrate the importance of second-hand smoke exposure as a risk factor for never-smoker lung cancer, but potentially not as the sole cause of LCINS.
Air pollution
Particulate matter air pollution with a diameter less than 2.5 μm (PM2.5) is classified as a carcinogen and a cause of lung cancer by the International Agency for Research on Cancer [59]. Recent analysis demonstrates a strong association between PM2.5 levels and lung cancer incidence [31,60,61]. Emerging data demonstrate increased mutation burden, including single-base substitution, double-base substitution, and indel mutations, as well as telomere shortening and increased incidence of TP53 mutations in patients diagnosed with lung cancer in areas with high PM2.5 levels [57]. Using genetically engineered mouse models, air pollution was found to drive an increase in macrophages and IL-1β release in the lung, fueling tumorigenesis [61], similar to the inflammatory process observed in patients with CHIP [48,49].
Identifying individuals at risk of lung cancer due to air pollution is challenging but could be achieved by classifying individuals by postcode based on air pollution levels. An alternative approach could involve targeting the inflammatory processes linked with air pollution in those individuals with high exposure. This approach is beginning to be explored in preclinical models of lung cancer [62]. However, lifetime exposure to air pollution is the sum of both indoor and outdoor exposures. Indoor air pollution (e.g., open fires, solid fuel stoves, poorly ventilated kitchens) is understudied, which makes risk stratification outside of urban environments, particularly in South Asia, parts of East and Southeast Asia, Africa, and some urban centers in the Middle East and South America, challenging.
Inflammatory conditions
Recent work has suggested an association between certain inflammatory conditions, such as gastroesophageal reflux disease, upper respiratory infections, type 1 diabetes mellitus, chronic gastrointestinal conditions, and viral infections, with an increased risk of lung cancer. However, the observational nature of these studies and potential confounding factors limit definitive conclusions about causality [36,63]. As for pulmonary inflammatory diseases, emerging evidence indicates that chronic obstructive pulmonary disease may represent a smoking-independent risk factor for lung cancer, whereas the association appears less definitive for alpha-1 antitrypsin deficiency [64,65].
Screening for early detection of LCINS
Detection of lung cancer at its earliest stages has a significant impact on patient mortality [66]. The National Lung Cancer Screening Trial (NLST), which used low-dose computed tomography (LDCT) to screen for lung cancer in high-risk individuals, demonstrated a 20% reduction in lung cancer-specific mortality [66,67]. High-risk individuals were defined as those aged 55–74 years with a smoking history of at least 30 pack-years [67]. Additional clinical trials of patients screened with LDCT have shown decreased lung cancer-related mortality, increased detection of early-stage lung cancer, and reduced late-stage disease [10,67,68]. However, without first selecting a high-risk population, these programs can cause unnecessary harm to participants, with false positives resulting in patient anxiety as well as unnecessary invasive procedures [66]. From an economic perspective, recent analyses suggest that LDCT screening can be cost-effective when targeted to high-risk groups [69]. Defining high-risk individuals in a never-smoker population is more difficult than in a smoker population, and the evidence of benefit from screening in never smokers is controversial [70,71].
Recent efforts to expand lung cancer screening to light- and never-smoker populations have proven complicated, as it is unclear if screening results in improved overall survival [71]. The recent Taiwan Lung Cancer Screening in Never-Smoker Trial (TALENT) highlighted the increased risk of lung cancer in never smokers with a family history of lung cancer in a first-degree relative [70]. As a result, the Taiwan National Screening Program now includes patients without a smoking history when there is a family history of lung cancer, with a 1.2 and 1.8% detection rate in males and females, respectively [70]. One systematic review and meta-analysis also demonstrated that lung cancer detection in female never smokers was just as effective as detection in male ever smokers in Asia [2]. These data together suggest that when a high-risk population is screened, in this case Asian female never smokers with a family history of lung cancer, LDCT can detect early-stage lung cancer and result in significant reductions in lung cancer-specific mortality. In an ongoing trial, Asian female never smokers based in New York (New York Female Asian Nonsmoker Screening Study) are being screened with LDCT. Preliminary data from this trial suggest that LDCT screening in this patient population is feasible, and the invasive adenocarcinoma detection rate is higher than in the NLST [72]. The data support further studies of whether populations with a family history of lung cancer could benefit from LDCT screening. Currently, there are no randomized trials to assess the survival and health economic benefit of LDCT in a never-smoker population.
Multi-cancer early detection testing is another approach for early lung cancer detection, with the aim to identify signals from multiple cancer types through cell-free DNA using a simple blood test [73–75]. So far, detection of early-stage disease has proven challenging [73,76]. These assays have shown great promise, even though tumor-informed panels seem to perform better in terms of tumor and recurrence detection [77,78].
Approaches for treatment and prevention of LCINS
Patients with lung cancer who have never smoked are a distinct patient population that requires a unique treatment strategy. Current guidelines for the treatment of patients diagnosed with lung cancer include assessment of driver mutations, stage, and grade, but patients are not stratified by smoking status. Features specific to lung cancer in never smokers, compared with smokers, include a higher proportion of adenocarcinoma histology, a significantly higher level of actionable driver mutations, a significantly lower mutational burden with unique mutational signatures, and a lower response rate and level of programmed death-ligand 1 (PD-L1) expression, a biomarker for ICI response in lung cancer [12,57,79–82] (Table 1). Classic immunotherapy treatments utilized in LCS have reduced efficacy in LCINS [81]. A combination of targeted therapies and immunotherapies that are specific to never-smoker lung cancer driver mutations, or alternative immunotherapy approaches involving neoantigen vaccination [83,84], may be more efficacious in this patient population.
