Lung cancer screening strategy for non-high-risk individuals: a narrative review.

Lung cancer is the deadliest malignancy worldwide, accounting for almost 20% of all cancer deaths. Clinical trials, such as NLST and NELSON, have proved the survival benefit of lung cancer screening using low-dose computed tomography (LDCT), and most of the lung cancer screening guidelines recommended annual lung cancer screening by LDCT for high-risk individuals. However, a relatively high proportion of lung cancer patients do not have risk factors, and it is questionable whether non-high-risk individuals should receive LDCT screening. In this review, we reviewed risk factors of lung cancer and summarized the benefits and potential harms of LDCT screening. After clarifying the differences between China and western countries in lung cancer screening, we recommended that non-high-risk individuals should receive LDCT screening with an interval of five to ten years. To better balance benefits and harms from LDCT screening, we also proposed a flexible screening strategy using LDCT based on lung cancer risk. Hopefully, it may help reduce unnecessary radiation exposure from CT scans while decreasing mortality of lung cancer. 2021 Translational Lung Cancer Research. All rights reserved.

Introduction

Lung cancer is the deadliest malignancy worldwide, with estimated over 1.7 million deaths in 2018 (1). Strenuous efforts have been made to improve the survival of lung cancer, and lung cancer screening is proved to be an effective method with relatively limited harm through early detection and treatment (2). Among all screening technologies, low-dose computed tomography (LDCT) stands out for its high sensitivity and non-invasiveness (3,4). The association between LDCT lung cancer screening and survival benefit was observed in National Lung Screening Trial (NLST) and Dutch-Belgian Lung Cancer Screening Trial (Nederlands-Leuvens Longkanker Screenings Onderzoek, NELSON) (5,6). However, NLST (5) enrolled asymptomatic participants aged 55–74 years and with a smoking history of at least 30 pack-year, and NELSON (7) enrolled individuals aged 50–75 years, who had smoked either ≥15 cigarettes per day for 25 years or ≥10 cigarettes per day for 30 years and were still smoking or had quit <10 years ago. These two trials only enrolled high-risk individuals who met the criteria of age and smoking histories, so most guidelines merely recommended LDCT lung cancer screening for high-risk individuals. Nevertheless, previous studies demonstrated that an estimated 40–60% of patients with lung cancer did not meet the US Preventative Task Force (USPTF) criteria, originated from the NLST eligibility criteria (8,9). Besides, 10% to 25% of all lung cancers occurred in never smokers (10,11). Therefore, non-high-risk individuals might also need lung cancer screening. This review aimed to give an overview on the risk factors of lung cancer, summarized benefits and harms of LDCT screening, and proposed a tailored screening programs using LDCT for non-high-risk individuals.

We present the following article in accordance with the Narrative Review reporting checklist (available at http://dx.doi.org/10.21037/tlcr-20-943).

Methods

The literature research was conducted in the PubMed/MEDLINE database on June 18th, 2020. The following search terms were used: lung cancer screening, low-dose CT screening, and lung nodule screening. Only English articles, published between January 1st, 1990 and June 18th, 2020, were enrolled in this review.

Discussion

Risk factors for lung cancer

In 2018, there were more than 2 million estimated new cases of lung cancer worldwide, accounting for 11.6% of all new cases with cancer (1). In the United States, 228 thousand individuals were estimated to develop lung cancer in 2020, accounting for 12.7% of all new cases with cancer (12). Especially in China, approximately 17.1% of new cancer cases had lung cancer (13). Therefore, to reduce the incidence of lung cancer, we need to identify risk factors.

Smoking

Cigarette smoking is considered to be a main risk factor. The association between the number of cigarettes and the risk of lung cancer was observed in previous studies (14,15). The second-hand smoking also contributed to the carcinogenesis of lung cancer (16). Smoking cessation could reduce lung cancer risk (16,17), and some guidelines also recommended smoking cessation while receiving lung cancer screening. Smoking addiction results from the presence of nicotine in tobacco. However, it is not nicotine but the exposure to tar (the total matter of smoke after removing nicotine and water) that leads to the carcinogenesis of lung cancer (18). Tar consists of 3,500 chemical substances and approximately 60 of them are known carcinogens (19).

Age

As time goes by, shortening of telomeres after many times of cell replication cycles and accumulative DNA damage may lead to the carcinogenesis of lung cancer. The young individuals have less possibility to develop lung cancer, compared with the old (20). However, the lung cancer incidence might be affected by the methods of lung cancer screening and the extent of ordinary persons understanding about lung cancer. Nowadays, with the development of the society, especially in some developing countries (such as China), LDCT is widely applied clinically, and the detection rate of lung cancer in young population increases (21,22).

Gender

Traditionally, men are more likely to develop lung cancer than women. However, the trends of lung cancer incidence in men and women changed dramatically. Jemal et al. reviewed the nationwide population-based incidence of lung cancer in America and found that the age-specific incidence decreased generally among both men and women with the age of 30 to 54 years old (23). Nevertheless, the declines among men have been steeper (23). During 2010 and 2014, lung cancer incidence was higher in women than men (23). In terms of mechanism, the expression of estrogen receptor (ER) alpha and ER beta were found to be increased in lung cancer tissues (24), and ER ligands could activate MAPK pathway and promote cell growth (25).

Environmental factors

Environmental factors may also contribute to carcinogenesis of lung cancer. They consist of indoor air pollution, including coal (26-28), biomass (29), and cooking fumes (30), outdoor air pollution (31,32), and occupational agents, including asbestos (33), arsenic (34), and silica (35,36). If we reduce the exposure to these environmental factors, the incidence of lung cancer should be decreased.

Genetic factors

Individuals response to the same environmental exposure differently. For instance, although smoking is considered to be a main risk factor of lung cancer, the majority of smokers will not develop lung cancer in their lifetime (37) and some never-smokers might also develop lung cancer. Moreover, the percentage of never smokers in lung cancer patients is about 10% to 20% in western countries (38), but it is as high as 50% to 63% in east-Asian population (39-41). Mechanically, gene nutation might be a significant factor and some genetic variations might contribute to the carcinogenesis of non-small cell lung cancer regardless of smoking history. In 2005, Bell et al. (42) reported a family with multiple cases of non-small cell lung cancer associated with the germline EGFR-T790M mutation. Subsequent studies identified relevant mutations in HER2, TP53, and BRCA2 as susceptible variations (43,44). Not only germline mutation but also single nucleotide polymorphism may result in the predisposition to lung cancer. Genome-wide association studies have identified rs3769821, rs2293607, rs1200399, rs17038564, rs35201538, and rs4573350 as risk loci for lung cancer in Chinese population (45).

Benefits of LDCT

Reducing lung cancer mortality

Reduction in lung cancer mortality remains as the major benefit for LDCT. In NLST (53,454 participants), there were 247 lung cancer-specific deaths per 100,000 person-years in the LDCT group and 309 lung cancer-specific deaths per 100,000 person-years in the radiography group, leading to a relative reduction in lung cancer mortality using LDCT of 20.0% [95% confidence interval (CI): 6.8–26.7%; P=0.004] (5). As for all-cause mortality, it was reduced in the LDCT group by 6.7% (95% CI: 1.2–13.6%; P=0.02) (5). Another lung cancer screening trial, NELSON (13,195 participants), investigated the association between LDCT lung cancer screening and survival benefit in 13,195 males and 2,594 females. Among men, lung cancer mortality was 2.50 deaths per 1,000 person-years for the LDCT group and 3.30 deaths per 1,000 person-years for the control group (no screening methods). The final results showed the cumulative rate ratio for lung cancer-specific death was 0.76 (95% CI: 0.61–0.94; P=0.01) at 10 years in the LDCT group, compared with the control group (6). Among women, the rate ratio was 0.67 (95% CI: 0.38–1.14) at 10 years (6). However, there are some conflicting results. The DLCST (46) (4,104 participants) reported a hazard ratio (HR) of 1.03 (95% CI: 0.66–1.60) in lung cancer mortality, the DANTE (47) (2,450 participants) reported a HR of 0.99 (95% CI: 0.69–1.43), and the MILD (48) (4,099 participants) reported a HR of 1.52 (95% CI: 0.63–3.65). The reason for no statistical difference in these trials may be the limited number of patients in the above studies, resulting in insufficient statistical power to detect lung cancer mortality reduction. Therefore, the results of NLST and NELSON are more convincing, and LDCT can reduce lung cancer mortality in high-risk individuals.

Other incidental findings besides lung cancer

During chest LDCT screening, abnormalities other than lung cancer could be found in lung, heart, chest wall, or breast, which might need clinical interference. In NLST, incidental findings were detected in 7.5% of participants (5,49), while they were detected in 6.7% of participants in NELSON (50). Although some unmeaningful findings may cause additional diagnostic processes and anxiety of patients, other clinically-relevant findings might lead to health benefits for participants. For instance, LDCT could be useful in the early detection of other smoking-related diseases besides lung cancer, such as COPD and cardiovascular diseases (51-53). Future studies are required to quantify the benefits of incidental findings on LDCT.

Harms of LDCT

Radiation exposure

The individuals receiving LDCT are exposed to radiation. The effective dose of radiation of LDCT is estimated to be 1.5 mSv per examination (54). One study based on the NLST data found that approximately one cancer death might be caused by radiation from imaging per 2,500 participants screened (54). According to the National Comprehensive Cancer Network (NCCN) lung cancer screening guidelines (55), annual screening LDCT is recommended for high-risk individuals. As a result, a single person needs to receive more than 20 times of LDCT in his or her lifetime. Therefore, to minimize the radiation risk from LDCT screening, the contradiction regarding expected benefits versus potential harms should be carefully balanced. Especially for the individuals with negative results at the baseline CT, the intervals of LDCT screening need to be carefully considered. In addition, less radiation dose per LDCT is expected due to revolutionized radiological technologies in the future.

False-positive findings

Some other lesions, such as inflammatory lesions, might showed a similar appearance as lung cancer. Most of the LDCT-detected lesions are not malignant, and the false-positive finding is an inevitable issue. In NLST, of all the participants in the three rounds of LDCT, 24.2% were classified as positive, and 23.3% were found to be false-positive, leading to the fact that 96.4% of the participants with positive results were false positive eventually (5). In NELSON, of those with positive results, 59.4% were considered to be false-positive, and the overall false-positive rate was 1.2% (7,56). The possible reason might be that the threshold for a positive screening test result in NELSON is higher than that in NLST (2). A proper CT scan strategy might help reduce the false-positive findings, and other methods to distinguish the malignant from the benign is also required in future studies.

Overdiagnosis and overtreatment

Overdiagnosis is defined as lung cancer detected by screening which would not otherwise affect the lifetime of patients if untreated, and overdiagnosis could lead to overtreatment. Overdiagnosis and overtreatment can result in unnecessary diagnostic procedures, the anxiety of patients, and increased medical expenses without any survival benefits. The excess analyses of NLST data suggested that the overdiagnosis rate in lung cancer patients was 18.5% (57), while the analyses based on microsimulation modeling demonstrated that was 9.6% (58). Therefore, the calculation of the overdiagnosis rate varies from the selected models. Besides, the analyses based on the data from clinical trials are flawed, because the enrolled participants must meet specific requirements and followed a relatively strict screening protocol. In real-world clinical practice, the scenarios could be much more complicated, so the results from clinical trials might be not as reliable as we thought. Moreover, the above analyses of overdiagnosis were based on the whole population. As for a single individual, it is not an easy task to identify whether he or she will be over-diagnosed, and he or she might be willing to reduce the possibility of lung cancer death by any means. Therefore, the overdiagnoses of LDCT might need a second thought in real-world clinical practice.

The differences between China and western countries in lung cancer screening

Most of lung cancer screening guidelines were released by western associations and societies (). NCCN recommended annual LDCT for either current smokers or former smokers quitting <15 years with a smoking history of ≥30 pack-year, who was 55 to 77 years old, or individuals with a smoking history of ≥20 pack-year and additional risk factors (other than second-hand smoke), who was no less than 50 years old (55). American Association for Thoracic Surgery suggested annual LDCT for individuals with a smoking history of ≥30 pack-year, who was 55 to 79 years old, or individuals with a smoking history of ≥20 pack-year and a cumulative risk of developing lung cancer of 5%, who was no less than 50 years old (59). For either current smokers or former smokers quitting <15 years with a smoking history of ≥30 pack-year, US Preventive Services Task Force recommended annual LDCT for individuals with the age between 55 and 80 (60), while American Cancer Society suggested it for those with the age between 55 and 74 (61). In addition, some risk prediction models have been developed (62-64). Based on data from the Beta-Carotene and Retinol Efficacy Study (CARET), Bach et al. (62) developed a lung cancer risk prediction model incorporating smoking history, asbestos exposure, sex, and age to predict annual absolute lung cancer risk of ever-smoking individuals aged over 45 years. Nevertheless, the current models were not representative for generalization and the selection of factors might be different in different models. Therefore, no prediction model is utilized clinically, and a good and unified prediction model is needed currently.

Table 1

The criteria for high-risk individuals in different guidelines

Institution	Age	Smoking history	Interval of LDCT
National Comprehensive Cancer Network (55)
Group 1	55–77	≥30 pack-year and either current smoker or former smoker quitting <15 years	Every year
Group 2	≥50	≥20 pack-year and with additional risk factors (other than second-hand smoke)	Every year
American Association for Thoracic Surgery (59)
Group 1	55–79	≥30 pack-year	Every year
Group 2	≥50	≥20 pack-year with a cumulative risk of developing lung cancer of 5%	Every year
USPSTF (60)	55–80	≥30 pack-year and either current smoker or former smoker quitting <15 years	Every year
American Cancer Society (61)	55–74	≥30 pack-year and either current smoker or former smoker quitting <15 years	Every year

The health insurance and access for LDCT were distinct between China and western countries. In China, LDCT is not only very cheap (about 30 US dollars), but also can be covered by healthcare insurance. The situation was completely different in America, where private health insurance and Medicare programs offer screening programs only for some eligible people. Moreover, in China, LDCT can usually be performed within the same day of appointment in some hospitals, which is not practical in European countries or America. As a result, in China, LDCT is easily accessible for both high-risk and non-high-risk individuals. Many employers in China have added LDCT into their regular annual health examination. As a result, almost everyone in China was accessible to LDCT if he or she wanted. Therefore, in clinical practice, some individuals from China received excessive LDCT, most of which was unnecessary.

Due to the discrepancies in clinical practice, the results of screening are different between China and western countries. Traditionally, smoking and age are considered to be risk factors for lung cancer (65,66). NLST and NELSON only enrolled high-risk individuals, who met the specific criteria on age and smoking history. The final results demonstrated that lung cancer screening using LDCT was associated with improved survival. Hence lung cancer screening guidelines, suggested that lung cancer screening using LDCT was only recommended for high-risk individuals annually (). The integration of smoking cession and lung cancer screening is recommended in many western countries (67). However, it is not the case in China. Previously, we analyzed LDCT screening data from 11,332 participants in Shanghai and found that there was no statistically significant difference between the incidences of primary lung cancer in smokers and nonsmokers (68). Furthermore, we investigated 8,392 employees from 6 hospitals in different regions of China, who received LDCT as a part of regular health examination (21). The results demonstrated that the lung cancer detection rate was significantly greater in female than male (2.5% vs. 1.3%, P=0.001). There was also a greater detection rate among nonsmokers than smokers, although there was no significant difference (2.2% vs. 1.4%, P=0.092). The detection rate in young employees was greater than before. Moreover, 95.5% of LDCT-detected lung cancer radiologically presented as ground-glass opacity (21). Another study, which investigated the lung cancer morbidity and mortality rate in Xuanwei City, China, also reveal high incidence in females and an early age peak in lung cancer death (22). Above studies revealed that the “traditionally-believed low-risk” population might also face a high risk of lung cancer. In spite of unknown reasons behind this phenomenon, this special group of individuals needs more attention and might benefit from lung cancer screening using LDCT.

The necessity of LDCT screening for non-high-risk individuals

In the United States, about 40–60% of lung cancer patients did not meet high-risk criteria of NLST (8,9). These patients should not receive LDCT according to the current guidelines of lung cancer screening. There are few existing clinical trials to investigate the use of LDCT in low-risk individuals. Most of evidence came from NLST (69), which was not originally designed for this. Wei et al. (70) investigated the performance of lung cancer screening with LDCT in 1,411 high-risk and 558 non-high-risk individuals in China. The results demonstrated that the positive rate was 9.7% for all participants and 11.3% for the high-risk individuals. In our clinical scenarios, a relatively high proportion of lung cancer patients did not have risk factors. Hence non-high-risk individuals might also need lung cancer screening. Moreover, Detterbeck (71) proposed that lung cancer could be divided into types with rapid growth, usual growth, slow growth, very slow growth, no growth and spontaneous regression according to their growth rates. LDCT screening intervention will inevitably detect more slow-growing tumors or indolent tumors especially for non-high-risk individuals. In fact, the majority of LDCT-detected lung cancers in non-high-risk individuals progress very slowly and are still at stage 0/I disease (21). Even if the non-high-risk participants develop new suspected lung cancer during the long screening interval, the situations are likely to be manageable and the survival should not be affected. Therefore, a longer interval will be more appropriate for non-high-risk individuals if the baseline CT is negative.

If we prolong the interval of LDCT to reduce the potential harms from it, LDCT might be beneficial for the non-high-risk individuals. We counted a 38-year-old woman with one lung nodule (), who did not have a history of smoking. She received lobectomy and systematic mediastinal lymph node dissection, and the pathological diagnosis was lung adenocarcinoma (pT2aN1M0). If this patient had received LDCT earlier, she may have just needed sublobar resection without lymph node dissection, and the prognosis would have been better. Moreover, we also met two 14-year-old teenagers without smoking history, whose pathological results were preinvasive adenocarcinoma (). Although it is not common in clinical practice, it emphasizes the importance of LDCT screening for non-high-risk individuals, especially for young and non-smoking individuals.

Figure 1

The computed tomography images for ground-glass opacity-featured lung adenocarcinoma. (A) The typical computed tomography image for a 38-year-old woman with lung cancer. (B,C) The typical computed tomography images for two 14-year-old teenagers with preinvasive lung cancer. Red arrows indicate lesions for patients with lung cancer.

The interval of LDCT screening for non-high-risk individuals

Currently, most guidelines on lung cancer screening recommend annual lung cancer screening using LDCT in high-risk individuals (). However, the choice of a yearly CT did not come from biological mechanisms, and it is debatable whether all eligible individuals should receive annual LDCT screening. Schreuder et al. (67) investigated the participants in NLST who underwent a baseline CT and a first annual negative follow-up scan and found out that the model, incorporating patient characteristics and baseline scan morphology, was significantly superior to the annual strategy. The Italian MILD trial remains as the only randomized controlled trial to compare different screening intervals using LDCT. In MILD (48), there were 4,099 high-risk participants, 1,723 randomized to the control group (no screening), 1,186 to the biennial LDCT group, and 1,190 to the annual LDCT group. No difference was observed in lung cancer mortality between the biennial and annual group. Biennial screening could save about one third of LDCT scans with similar performance indicators as compared to annual screening (72). Therefore, we believe that two years is a better interval for high-risk individuals using LDCT screening, and future trials are urged to clarify this issue. As for non-high-risk individuals, an interval of more than two years has been suggested for low-risk individuals (73,74). Therefore, we believe that the current strategy for lung cancer screening using LDCT might be too tight, and a loose strategy should be more appropriate.

Here, we proposed a new screening strategy with flexible intervals of LDCT based on lung cancer risks (). For participants with an age of less than 50, the interval might be five to ten years, if the baseline CT is negative. For those between 50 and 60 years old, the interval might be three to five years according to the number of risk factors. For those between 60 and 70 years old, it might be two to three years. For those with an age of more than 70 years, it might be two years.

Figure 2

A screening strategy for lung cancer screening based on risk factors.

Conclusions

In summary, there are some differences in LDCT screening between China and western countries. LDCT should be recommended for not only high-risk but also non-high-risk individuals. To balance the benefits and harms, the non-high-risk individuals should receive LDCT every five or ten years. We also put forward a screening strategy based on lung cancer risk. Hopefully, it might help reduce radiation exposure from CT scans while decreasing the mortality of lung cancer.

The article’s supplementary files as