Chemoprevention of lung cancer

INTRODUCTION — Lung cancer is the leading cause of cancer deaths worldwide, with an estimated 2 million new cases and over 1.8 million deaths in 2018 [1]. Tobacco smoking is responsible for most cases of lung cancer (approximately 90 percent for men and 70 to 85 percent for women). The risk of lung cancer with chronic marijuana smoking and use of electronic nicotine delivery devices remains to be determined. (See "Cigarette smoking and other possible risk factors for lung cancer".)

Never smoking or smoking cessation for smokers are the only proven means to decrease the risk of developing lung cancer. Former smokers continue to have an elevated risk of developing lung cancer for at least 30 years after stopping smoking, making these individuals an important target group for further efforts at mortality reduction [2,3]. (See "Overview of smoking cessation management in adults" and "Screening for lung cancer" and "Benefits and consequences of smoking cessation" and "Pharmacotherapy for smoking cessation in adults".)

Chemoprevention is the use of dietary or pharmacologic agents to prevent or slow the progression of cancer [4]. However, there is no convincing evidence that any approach other than smoking cessation can decrease the risk of lung cancer.

Multiple agents have been studied to decrease the incidence of lung cancer, particularly in those at high risk for the development of this disease. The rationale for various approaches to chemoprevention (beyond never smoking or smoking cessation), observational data and their implications for epidemiologic studies, and results of chemoprevention trials that have been conducted are discussed in this topic.

IDENTIFYING PATIENTS AT HIGH RISK FOR LUNG CANCER — For chemoprevention to be feasible, a high-risk population must be identified and an effective chemopreventive agent with minimal side effects must be available. Current or former smokers with an annual risk of up to 2 percent are identifiable using a combination of clinically available risk factors, including: smoking history, age, sex, presence of airflow obstruction or emphysema, environmental/occupational exposure, a history of a tobacco-related aerodigestive cancer, and family history of lung cancer [5,6].

The role of radiographic screening for lung cancer in patients at high risk is discussed elsewhere. (See "Screening for lung cancer".)

IS THERE A TARGETABLE GENETIC LESION? — All of the histologic cell types of lung cancer are genetically complex, with squamous cell demonstrating the highest frequency of mutations [7]. No common mutation is shared across a majority of lung cancers, making a targeted approach to chemoprevention challenging. However, common pathways for the early stages of premalignancy (pulmonary inflammation, tissue hypoxia, aging) or distinct phenotypes susceptible to lung carcinogenesis may be definable and targeted for intervention [8-10].

Process of carcinogenesis — Carcinogenesis in the lung proceeds through a multistep process [11]. The airways of smokers are directly exposed to a complex mixture of carcinogens in tobacco smoke, which can result in a field effect with multiple premalignant clonal lesions [12,13]. Invasive tumors are often surrounded by an area of relatively normal epithelium harboring some of the same mutations that are found in the carcinoma, supporting the multistep carcinogenesis hypothesis [14]. Whole-genome sequencing of cultured clones arising from single bronchial epithelial cells from a series of 16 individuals of differing age and smoking history has revealed that tobacco smoking has a massive effect on the mutational burden of these cells, with 1000 to 10,000 mutations per cell in current smokers [15]. In former smokers, a population of bronchial epithelial cells with a much lower mutational burden emerged, demonstrating expansion of a population of mitotically quiescent progenitor cells that could presumably repair tobacco-damaged epithelium.

●A central airway progression of premalignant lesions (squamous metaplasia; mild, moderate, and severe dysplasia; and carcinoma in situ) has been described, which leads to invasive squamous cell carcinoma [16]. A subset of these premalignant lesions exhibits gene copy number alterations and is associated with increased squamous cell lung cancer risk [17,18].

These premalignant lesions may be sampled either by routine white light or autofluorescence bronchoscopy [19]. Histologic changes at a single timepoint, with the possible exception of carcinoma in situ, are not highly predictive of the future development of invasive squamous cell lung cancer [20]. The evolution of a dysplastic lesion, regression or persistence and progression, is significantly associated with risk of incident squamous cell lung cancer, although these often arise at a site distinct from the persistent/progressive lesion [21]. Carcinoma in situ is also often identified adjacent to invasive squamous cell carcinoma. The finding of carcinoma in situ in the absence of carcinoma should usually prompt a careful evaluation, including computed tomography (CT) and repeat bronchoscopy with multiple biopsies, to determine if an occult invasive lesion is present.

A comprehensive genomic, epigenomic, and transcriptomic analysis of carcinoma in situ lesions that have either progressed to invasive squamous cell carcinoma or regressed has yielded important insights into mechanisms driving progression [22]. One consistent signal for progression is expression of genes associated with chromosomal instability, including expression of the genes ACTL6A, ELAVL1, MAD2L1, NEK2, and OIP5.

●The immune microenvironment is associated with progression or regression of squamous cell premalignancy, supporting the concept that regression is often an immune-mediated process [23-25].

●An analogous, but histologically distinct, series of premalignant lesions, including atypical adenomatous hyperplasia and adenocarcinoma in situ, occurs in the peripheral lung; these lesions can progress to invasive adenocarcinoma [26]. Many of these lesions are initially detected on low-dose chest CT scans performed for lung cancer screening, or they are incidentally noted on chest CT scans done for other clinical indications. (See "Pathology of lung malignancies", section on '2021 classification of adenocarcinoma'.)

The understanding of premalignant genetic alterations in peripheral lung lesions that give rise to adenocarcinoma has been hampered by difficulty in obtaining biospecimens other than in surgical resections, although new bronchoscopic techniques may alleviate this partially. Serial sampling of peripheral premalignancy to understand mechanisms of regression remains in its infancy. Premalignant lesions separate from a resected adenocarcinoma have undergone genomic investigation [27]. Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations were identified in a minority (18 percent) of smoking-associated lesions, but were not seen in lesions arising in never-smokers. B-Raf proto-oncogene (BRAF) mutations were discovered in 23 percent of patients, both in ever- and never-smokers. Gene set expression analysis supported the hypothesis that antitumor immune signaling is suppressed and protumor immune signaling is enhanced in atypical adenomatous hyperplasia.

Peripheral premalignant lesions are often identifiable as small ground-glass infiltrates on CT scans, although ground-glass infiltrates are nonspecific and can arise from inflammation as well as premalignancy [28]. With the widespread use of CT screening for lung cancer, more atypical adenomatous hyperplasia lesions will likely be identified. (See "High resolution computed tomography of the lungs", section on 'Ground-glass opacification'.)

●Premalignant lesions for small cell lung cancer have not been described.

INVESTIGATIVE STRATEGIES — Epidemiologic and observational studies have established the importance of smoking cessation, but have not led to other effective lung cancer prevention strategies. The health effects of smoking cessation are addressed elsewhere. (See "Benefits and consequences of smoking cessation", section on 'Malignancy'.)

Data regarding three agents (inhaled corticosteroids, low-dose aspirin, anti-interleukin 1 beta [IL-1 beta] monoclonal antibodies) have attracted interest as possible agents for lung cancer chemoprevention. However, multiple phase III lung cancer chemoprevention trials have failed to establish the efficacy of a variety of approaches, which emphasizes the fact that current approaches to predict efficacy are inadequate [29]. We suggest not using these agents for chemoprevention against lung cancer, given that further evaluation in randomized trials is required.

Epidemiologic studies have also identified a number of factors that appear to be related to a decreased incidence of lung cancer. However, it is likely that many of these associations are confounded by covarying factors, such as socioeconomic status, education, and smoking.

Clinical settings — Chemoprevention is classified as primary, secondary, or tertiary, based upon the target population:

●Primary chemoprevention – Primary chemoprevention refers to preventing cancer in healthy individuals who are at increased risk. This population primarily includes current and former smokers, as well as those with exposure to known lung carcinogens such as asbestos and radon.

●Secondary chemoprevention – Secondary chemoprevention focuses on blocking the development of cancer in individuals in whom a precancerous lesion has been detected. (See 'Is there a targetable genetic lesion?' above.)

●Tertiary chemoprevention – Tertiary chemoprevention targets patients who had a previous lung or other tobacco-related cancer in an effort to prevent the development of a second primary tumor.

Primary prevention

Aspirin — Although aspirin has demonstrated a protective benefit on lung cancer deaths in a randomized trial, it remains unclear whether this is from suppression of metastases in patients who already have lung cancer versus reduced incidence of new cancers. Observational data support a chemopreventive role for aspirin. Further randomized trial data are needed before using aspirin for lung cancer prevention.

●Low-dose aspirin has been extensively studied for the prevention of cardiovascular disease. Pooled data from seven such trials in which multiple individual patient outcomes were available have been correlated with both cancer deaths and cancer metastasis [30]. The analysis included 23,535 patients with 657 cancer deaths, of which 198 were from lung cancer. Benefit in cancer deaths was only seen after five years' or more follow-up; deaths from lung cancer showed a consistent reduced hazard ratio (HR) across various time periods ranging from 0.68 to 0.75, with statistically significant values at 0 to 10 years and 0 to 20 years' follow-up. The potential benefit was only observed for adenocarcinoma, not other cell types. A further analysis of this study showed similar reductions in lung cancer metastasis, suggesting that at least some of the effect may have been due to suppression of metastasis, rather than interference in earlier stages of carcinogenesis [31].

However, in a retrospective cohort study of almost 13 million participants in the Korean National Health Information Database, intake of low-dose aspirin (100 mg or less) for at least five years was associated with a modestly reduced risk of lung cancer (eg, compared with no aspirin use; HR 0.96 with five to six years of aspirin use and reaching 0.89 with nine years of aspirin use) [32]. After stratified analysis, a more pronounced reduction of lung cancer risk was observed among people aged 65 years or older and among people without diabetes.

●Results from the Women's Health Study also support to a possible chemopreventive activity of low-dose aspirin for lung cancer. In the Women's Health Study, a 2 x 2 design study of low-dose aspirin and vitamin E, newly diagnosed invasive cancer was the primary endpoint, with incident breast, colorectal, and lung cancer as predetermined secondary endpoints [33]. No effect of aspirin on total incident cancer or incident breast or colorectal cancer was found, with or without vitamin E. However, a trend toward a decreased incidence of lung cancer was noted (relative risk [RR] 0.78, 95% CI 0.59-1.03), and there was a statistically significant decrease in lung cancer mortality (RR 0.7, 95% CI 0.50-0.99; p = 0.04). Although encouraging, these results must be tempered by consideration of multiple endpoint testing and a potential effect on lung cancer death through suppression of metastasis rather than incidence.

Vitamins and minerals — Diets high in fruit and vegetables have been consistently associated with reduced lung cancer risk [34]. However, no positive trials of dietary intervention to decrease the risk of lung cancer have been reported. Reduced serum levels of antioxidants and vitamins, particularly vitamin A, have been reported in patients with lung cancer, further supporting the concept that supplementation might be protective. Unfortunately, randomized trials of chemoprevention with vitamins and antioxidants have shown no benefits in regards to reduced lung cancer, and these strategies are not used as chemoprevention.

●Vitamin E and beta carotene – Three primary chemoprevention trials have evaluated: the combination of vitamin E and beta carotene (the alpha tocopherol beta carotene or ATBC study), beta carotene and retinol (CARET), or beta carotene alone [35-37]. None showed benefit.

•In regards to beta carotene, both CARET and ATBC demonstrated statistically significant increases (RR 1.28 and 1.18, respectively) in lung cancer risk in the treatment groups with this vitamin.

•In regards to vitamin E, a benefit was not observed.

●A combination of multivitamins and minerals was ineffective in a large primary prevention trial [38].

Inhaled corticosteroids — Pulmonary inflammation induced by tobacco smoke may contribute to the development of lung cancer, but further data are required before inhaled corticosteroids can be supported as a chemoprevention strategy for lung cancer.

The effect of inhaled corticosteroids on the natural history of chronic obstructive pulmonary disease has been studied in several double-blinded, randomized trials. Although no reduction in lung cancer incidence was observed, the mean duration of these trials was only 26 months, which may have been too short to demonstrate a chemopreventive effect.

By contrast, an analysis of patients at Veterans Affairs Medical Centers who were prescribed inhaled corticosteroids for chronic obstructive pulmonary disease and were highly compliant revealed a significantly reduced risk of lung cancer [39].

Canakinumab (IL-1 beta inhibitor) — The monoclonal antibody canakinumab targets interleukin 1 beta (IL-1 beta) and is an investigative option, being evaluated for several conditions, but is not used for lung cancer prevention outside of a clinical trial.

Canakinumab was studied in a randomized, placebo-controlled trial with the primary endpoint of reducing vascular events in patients with a previous myocardial infarction and an elevated level of high-sensitivity C-reactive protein (2 mg/L or greater). Lung cancer incidence was a prespecified secondary endpoint and was reduced significantly in the canakinumab group in a dose-dependent fashion ranging from HRs of 0.61 to 0.33 over a median follow-up period of 3.7 years [40]. The short follow-up period makes it unlikely that canakinumab inhibits early events in lung carcinogenesis, but more likely affects progression or metastasis. Overall mortality was not improved, largely due to an increase in death from sepsis. Canakinumab is being studied in many early-stage lung cancer treatment trials, and several of these trials are collecting second primary tumor data.

Secondary chemoprevention — Secondary chemoprevention refers to the prevention of progression to lung cancer for people who have evidence of premalignancy. Several agents, including oral iloprost and celecoxib, have achieved promising results in phase II trials as secondary prevention, but the endpoints that were modulated have not been validated as predictive. As such, further data are required prior to use of these agents as secondary chemoprevention against lung cancer.

Ideally, surrogate endpoints for lung cancer chemoprevention should be:

●Associated with lung cancer

●Mechanistically involved in the carcinogenesis process

However, studies to date have used surrogate endpoints including endobronchial dysplasia, bronchial epithelial proliferation as measured by Ki-67 immunostaining, and the presence of computed tomography (CT)-detected lung nodules. None of these endpoints have been validated as fulfilling the above criteria of an ideal surrogate endpoint and are justified on the basis of biologic plausibility.

A number of studies have used bronchoscopy to target premalignant airway lesions but few of these have met their primary endpoints. Examples include the following [41]:

●Negative results have been reported for 13-cis retinoic acid [42,43], fenretinide [44], etretinate [45], beta carotene [46], vitamin B12 and folate [47], and inhaled corticosteroids [48,49].

●A study of anethole dithiolethione, a compound found in green tea, did not achieve its primary endpoint of reducing new dysplastic lesions but did decrease the rate of worsening of endobronchial lesions, a secondary endpoint [50].

●Two trials of the COX-2 inhibitor celecoxib with bronchial epithelial proliferation as the primary endpoint have been positive [51,52]. The biologic basis for COX-2 inhibition is attractive. COX-2 overexpression is seen in non-small cell lung cancer (NSCLC) and adenocarcinoma precursor lesions, and COX-2 inhibition inhibits chemical carcinogenesis in murine lung cancer models.

●One trial of the oral prostacyclin analog iloprost compared with placebo demonstrated a statistically significant improvement in the primary endpoint, bronchial dysplasia, but only in former smokers [53]. The improvement in bronchial dysplasia was similar to the difference between current and former smokers. Oral iloprost is no longer available, but a phase I trial of inhaled iloprost has been completed.

In preclinical models, prostacyclin and iloprost do not act through the canonical cell surface prostacyclin receptor, but rather at least partially through peroxisomal proliferator-activated receptor (PPAR) gamma activation [54]. A phase II trial of oral pioglitazone (a PPAR gamma activator) with the endpoint of bronchial dysplasia was completed and did not show evidence of efficacy [55], suggesting that differences in the mechanisms of activity exist.

Several trials have assessed innovative endpoints:

●In one trial of inhaled budesonide with the primary endpoint of bronchial dysplasia, dysplasia did not improve, but a secondary endpoint, the proportion of CT-detected pulmonary nodules, did respond [56]. Two subsequent phase II trials of inhaled corticosteroid with the primary endpoint of CT-detected pulmonary nodules were negative, however [48,49]. Although CT-detected nodules (particularly ground-glass nodules) are a nonspecific endpoint, some do represent adenocarcinoma premalignancy, and further consideration of using this novel surrogate endpoint, perhaps with histologic confirmation, is warranted.

●Gene expression analysis of endobronchial brushings has been carried out in a small phase I trial of myo-inositol, with the finding that activation of the phosphatidylinositol 3-kinase (PI3K) pathway was associated with response to myo-inositol, which has PI3K inhibitory effects [57]. A follow-up phase II randomized placebo-controlled trial failed to demonstrate either efficacy of myo-inositol or activation of the PI3K pathway as a predictive biomarker [58]. While disappointing, these two trials have raised enthusiasm for studying alterations in premalignancy and precision chemoprevention where agents target specific activated pathways [59].

Prevention trials have also started to evaluate modulation of the immune response. Immunoprevention trials evaluating checkpoint inhibitors are recruiting (nivolumab for premalignant squamous cell lesions [NCT03347838], and a trial of pembrolizumab in pulmonary nodules [NCT03634241]).

Tertiary prevention — No interventions have reduced the occurrence of a second or recurrent lung cancer in patients previously diagnosed with lung cancer, with the exception of smoking cessation.

In a 2010 meta-analysis of 10 observational studies including patients with early-stage lung cancer, continued smoking was associated with increased all-cause mortality (HR 2.9) and recurrence (HR 1.9) in early-stage NSCLC, and in all-cause mortality (HR 1.9), development of a secondary primary tumor (HR 4.3), and recurrence (HR 1.3) in limited small cell lung cancer [60]. No study contained data on the effect of quitting smoking on cancer-specific mortality or the development of a second primary tumor in NSCLC.

Subsequent studies have confirmed these findings [61-63]. As an example, among over 7059 patients diagnosed with primary lung cancer, 2.3 percent developed a second primary lung cancer [61]. Smoking cessation was associated with an 83 percent RR reduction for second primary lung cancer (HR 0.17).

Further discussion on the health benefits of smoking cessation is found elsewhere. (See "Benefits and consequences of smoking cessation", section on 'Malignancy'.)

Examples of available data regarding other interventions include the following:

●Selenium supplementation was associated with decreased incidence of lung cancer in a trial with skin cancer as the primary endpoint but was ineffective in a subsequent tertiary prevention trial with second primary lung cancer as the endpoint [64].

●13-cis retinoic acid and the combination of vitamin A and n-acetyl cysteine have been evaluated in two trials of patients with a history of cancer (tertiary chemoprevention), both of which showed no beneficial effect [65,66].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diagnosis and management of lung cancer".)

SUMMARY

●Introduction – Never-smoking and smoking cessation among smokers clearly reduce the risk of lung cancer, and the leading prevention effort for current smokers should be smoking cessation. Chemoprevention is the use of dietary or pharmacologic agents to prevent or slow the progression of cancer [4]. However, there is no convincing evidence that any approach other than smoking cessation can decrease the risk of lung cancer. (See 'Identifying patients at high risk for lung cancer' above.)

●Investigative strategies – Although lung cancer chemoprevention has considerable potential, the strategies discussed below are investigational and we do not use them for lung cancer chemoprevention.

•Primary prevention – Primary chemoprevention refers to preventing cancer in healthy individuals who are at increased risk. This population primarily includes current and former smokers, as well as those with exposure to known lung carcinogens.

Efforts to prevent lung cancer have included a variety of agents, including aspirin, beta carotene, multivitamins, and the interleukin 1 beta inhibitor canakinumab. None have been successful in phase III trials, and beta carotene supplementation actually increases lung cancer risk, emphasizing that the administration of supplements with a good biologic rationale to reduce risk is not necessarily benign. (See 'Investigative strategies' above.)

•Secondary prevention – Secondary chemoprevention focuses on blocking the development of cancer in individuals in whom a precancerous lesion has been detected.

Several agents, including oral iloprost and celecoxib, have achieved promising results in phase II trials, but the endpoints that were modulated have not been validated as predictive.

•Tertiary prevention – No interventions have reduced the occurrence of a second or recurrent lung cancer in patients previously diagnosed with lung cancer, with the possible exception of smoking cessation. Selenium supplementation was associated with decreased incidence of lung cancer in a trial with skin cancer as the primary endpoint but was ineffective in a subsequent tertiary prevention trial with second primary lung cancer as the endpoint.

ACKNOWLEDGMENT — The editorial staff at UpToDate acknowledge Arthur Skarin, MD and Ravi Salgia, MD, PhD, who contributed to an earlier version of this topic review.