Abstract
This study assesses non–risk-based low-dose computed tomographic screening to assess lung cancer detection in the general population and evaluate outcomes by risk stratification.
Low-dose computed tomography (LDCT) is an effective lung cancer (LC) screening tool endorsed by the National Comprehensive Cancer Network (NCCN) for high-risk individuals and based on randomized clinical trials in older and smoking populations.1 However, studies indicate increasing LC cases among younger, nonsmoking populations, particularly in Asia.2,3 Based on China’s LC screening consensus,4 a non–risk-based LDCT screening was conducted to assess LC detection in the general population and evaluate outcomes by risk stratification.
Methods
Guangzhou Lung-Care Project, a prospective intervention cohort study, enrolled participants aged 40 to 74 years from December 1, 2015, to July 31, 2021. Exclusions included prior LC diagnosis or treatment within 5 years, chest CT within the past year, or significant cancer-related symptoms. Primary outcomes were LC detection rates and stage I proportion. LDCT diagnostic performance was assessed, with outcomes stratified by NCCN and Chinese consensus criteria.1,4 Recruitment, screening, diagnostic evaluation, and follow-up are detailed in the eMethods in Supplement 1. Data were censored in August 2022. Statistical analyses used IBM SPSS version 24.0. The First Affiliated Hospital of Guangzhou Medical University institutional review board approved the study. Participants provided written informed consent.
Results
Among 11 708 participants (median age, 59 years; IQR, 51-65), 1883 (16.1%) and 4902 (41.9%) cases were classified as high risk for LDCT screening using NCCN and Chinese consensus criteria, respectively (Table 1). Positive results were detected in 2245 participants (19.2%). Among 231 participants who underwent diagnostic examinations, 200 participants (1.7%) received an LC diagnosis via biopsies, with 82.5% at stage I; 20 biopsy results were benign and 10 revealed nonlung malignancies. Repeat CT scans decreased from 908 (7.8%, second scan) to 5 (0.04%, seventh scan). Biopsy complications occurred in 25.5% of participants (59 of 231), including 21.6% (50 of 231) minor to intermediate events and 3.9% (9 of 231) major events. No procedural deaths occurred within 60 days. Onetime LDCT diagnostic performance is detailed in Table 2. The median follow-up was 6.7 years (IQR, 6.2-7.4 years). By September 2024, 28 patients had died from LC, predominantly stage II to IV disease. Seven interval LC cases were found during follow-up. Only 38 and 112 of diagnosed LCs met criteria per NCCN and Chinese consensus, respectively. The missed diagnosis rates for NCCN and Chinese consensus were 81.0% and 44.0%, respectively. LC detection rates among high-risk and non–high-risk populations were 2.0% and 1.6% per NCCN and 2.3% and 1.3% per Chinese consensus. Similar trends were observed in sensitivity analyses by excluding cases of adenocarcinoma in situ, minimally invasive adenocarcinoma, or both when defining LC (Table 2). In addition, the proportions of stage I diseases among high-risk and non–high-risk populations were 60.5% and 92.0% per NCCN criteria and 80.4% and 93.2% per Chinese consensus.
Table 1. Clinical and Diagnosis Characteristics of Participants in Screening.
| Characteristic | Total, No. (%) (N = 11 708) |
|---|---|
| Sex | |
| Male | 5452 (46.6) |
| Female | 6256 (53.4) |
| Age, y | |
| 40-44 | 1047 (8.9) |
| 45-49 | 1361 (11.6) |
| 50-54 | 1636 (14.0) |
| 55-59 | 2022 (17.3) |
| 60-64 | 2469 (21.1) |
| 65-69 | 2118 (18.1) |
| 70-74 | 1055 (9.0) |
| Smoking status | |
| Current smoker, pack-years | 2070 (17.7) |
| <10 | 232 (2.0) |
| 10-19 | 385 (3.3) |
| ≥20 | 1453 (12.4) |
| Previously smoked | 1063 (9.1) |
| Smoking cessation <15 y | 410 (3.5) |
| Smoking cessation ≥15 y | 653 (5.6) |
| Never smoked | 8102 (69.2) |
| Missing | 473 (4.0) |
| Family history of cancer | |
| Yes | 3122 (26.7) |
| Lung | 1158 (9.9) |
| Other | 1964 (16.8) |
| No | 8327 (71.1) |
| Missing | 259 (2.2) |
| History of cancer | |
| Yes | 358 (3.1) |
| No | 11 283 (96.4) |
| Missing | 67 (0.6) |
| History of chronic respiratory disease | |
| Yes | 1435 (12.3) |
| COPD or emphysema | 906 (7.7) |
| Pulmonary fibrosis | 68 (0.6) |
| Pulmonary tuberculosis | 616 (5.3) |
| No | 10 273 (87.7) |
| Missing | 0 |
| Occupational exposure to carcinogens | |
| Yes | 3129 (26.7) |
| Asbestos | 93 (0.8) |
| Oil paint | 747 (6.4) |
| Coal dusta | 1731 (14.8) |
| Radon | 43 (0.4) |
| Rubber | 56 (0.5) |
| Pesticide | 166 (1.4) |
| Radiation | 72 (0.6) |
| Arsenic | 26 (0.2) |
| Silicon dioxide | 19 (0.2) |
| Heavy metals | 285 (2.4) |
| No | 8551 (73.0) |
| Missing | 18 (0.2) |
| History of asthma | |
| Yes | 448 (3.8) |
| No | 11 260 (96.2) |
| Food allergy | |
| Yes | 291 (2.5) |
| No | 11 417 (97.5) |
| Allergy to temperature change | |
| Yes | 1932 (16.5) |
| No | 9776 (83.5) |
| Classified as high risk and non–high risk | |
| High risk using NCCN criteria | 1883 (16.1) |
| Non–high risk using NCCN criteria | 9825 (83.9) |
| High risk using Chinese consensus | 4902 (41.9) |
| Non–high risk using Chinese consensus | 6806 (58.1) |
Abbreviations: COPD, chronic obstructive pulmonary disease; NCCN, National Comprehensive Cancer Network.
Individuals who used coal or coal gas for cooking (1619 of 1731) were included as a part of the population exposed to coal.
Table 2. Overall Diagnostic Performance of Onetime LDCT Stratified by Different Guidelines.
| Stratification | Lung cancer detection rate, % | Sensitivity, % | Specificity, % | PPV, % | NPV, % | Accuracy, % |
|---|---|---|---|---|---|---|
| Nonrisk based (current study) | 1.7 (200 of 11 708) | 96.6 (200 of 207) | 82.2 (9456 of 11 501) | 8.9 (200 of 2245) | 99.9 (9456 of 9463) | 82.5 (9656 of 11 708) |
| High-risk individuals in NCCN | ||||||
| All | 2.0 (38 of 1883) | 90.5 (38 of 42) | 81.9 (1508 of 1841) | 10.2 (38 of 371) | 99.7 (1508 of 1512) | 82.1 (1546 of 1883) |
| Excluding AIS | 2.0 (38 of 1883) | 90.5 (38 of 42) | 81.9 (1508 of 1841) | 10.2 (38 of 371) | 99.7 (1508 of 1512) | 82.1 (1546 of 1883) |
| Excluding AIS and MIA | 1.9 (35 of 1880) | 89.7 (35 of 39) | 81.9 (1508 of 1841) | 9.5 (35 of 368) | 99.7 (1508 of 1512) | 81.9 (1543 of 1880) |
| Low-risk individuals in NCCN | ||||||
| All | 1.6 (162 of 9825) | 98.2 (162 of 165) | 82.3 (7948 of 9660) | 8.6 (162 of 1874) | 100 (7948 of 7951) | 82.5 (8110 of 9825) |
| Excluding AIS | 1.6 (155 of 9818) | 98.1 (155 of 158) | 82.3 (7948 of 9660) | 8.3 (155 of 1867) | 100 (7948 of 7951) | 82.5 (8103 of 9818) |
| Excluding AIS and MIA | 1.2 (115 of 9778) | 97.5 (115 of 118) | 82.3 (7948 of 9660) | 6.3 (115 of 1827) | 100 (7948 of 7951) | 82.5 (8063 of 9778) |
| High-risk individuals in Chinese consensus | ||||||
| All | 2.3 (112 of 4902) | 94.9 (112 of 118) | 81.5 (3899 of 4784) | 11.2 (112 of 997) | 99.8 (3899 of 3905) | 81.8 (4011 of 4902) |
| Excluding AIS | 2.2 (109 of 4899) | 94.8 (109 of 115) | 81.5 (3899 of 4784) | 11.0 (109 of 994) | 99.8 (3899 of 3905) | 81.8 (4008 of 4899) |
| Excluding AIS and MIA | 1.9 (93 of 4883) | 93.9 (93 of 99) | 81.5 (3899 of 4784) | 9.5 (93 of 978) | 99.8 (3899 of 3905) | 81.8 (3992 of 4883) |
| Low-risk individuals in Chinese consensus | ||||||
| All | 1.3 (88 of 6806) | 98.9 (88 of 89) | 82.7 (5557 of 6717) | 7.1 (88 of 1248) | 100 (5557 of 5558) | 82.9 (5645 of 6806) |
| Excluding AIS | 1.2 (84 of 6802) | 98.8 (84 of 85) | 82.7 (5557 of 6717) | 6.8 (84 of 1244) | 100 (5557 of 5558) | 82.9 (5641 of 6802) |
| Excluding AIS and MIA | 0.8 (57 of 6775) | 98.3 (57 of 58) | 82.7 (5557 of 6717) | 4.7 (57 of 1217) | 100 (5557 of 5558) | 82.9 (5614 of 6775) |
Abbreviations: AIS, adenocarcinoma in situ; LDCT, low-dose computed tomography; MIA, minimally invasive adenocarcinoma; NCCN, National Comprehensive Cancer Network; NPV, negative predictive value; PPV, positive predictive value.
Discussion
Previous LC screening mainly targeted smokers.5 With increasing LC incidence among nonsmokers in East Asia, Chang et al2 conducted LDCT screening in never-smokers, finding a 2.6% detection rate. Our study expanded inclusion criteria beyond preestablished risk factors, yielding comparable detection rates. We found that the LC detection rate among individuals classified as not high risk was comparable to that of high-risk individuals, with a higher proportion of stage I cancer detected in the group without high risk. Although only 7 cases were adenocarcinoma in situ, potential overdiagnosis concerns remain, given 21.5% minimally invasive adenocarcinoma. Future research should focus on identifying the characteristics of LC that truly require therapeutic intervention to alter their prognosis.6 This study is limited by its single-group, unrandomized design without mortality data from an unscreened comparison group, which prevents the assessment of the value of screening this population. It is essential to test the generalizability of these findings in other regions and races (eg, the US). This study revealed that LDCT has a considerable detection rate of LCs in populations without known high-risk factors. These findings highlight the necessity of prospective studies to evaluate efficacy of CT screening and the importance of identifying high-risk factors or prescreen enriching biomarkers in populations not traditionally considered high risk.
eMethods. Supplemental Methods
eReferences.
Data Sharing Statement
References
- 1.Wood DE, Kazerooni EA, Aberle D, et al. NCCN Guidelines insights. J Natl Compr Canc Netw. 2022;20(7):754-764. doi: 10.6004/jnccn.2022.0036 [DOI] [PubMed] [Google Scholar]
- 2.Chang GC, Chiu CH, Yu CJ, et al. ; TALENT Investigators . Low-dose CT screening among never-smokers with or without a family history of lung cancer in Taiwan. Lancet Respir Med. 2024;12(2):141-152. doi: 10.1016/S2213-2600(23)00338-7 [DOI] [PubMed] [Google Scholar]
- 3.Yang W, Qian F, Teng J, et al. ; Written on behalf of the AME Thoracic Surgery Collaborative Group . Community-based lung cancer screening with low-dose CT in China. Lung Cancer. 2018;117(117):20-26. doi: 10.1016/j.lungcan.2018.01.003 [DOI] [PubMed] [Google Scholar]
- 4.Chinese Thoracic Society, Chinese Medical Association; Chinese Alliance Against Lung Cancer Expert Group . Chinese expert consensus on diagnosis and treatment of pulmonary nodules (2024). Article in Chinese. Zhonghua Jie He He Hu Xi Za Zhi. 2024;47(8):716-729. doi: 10.3760/cma.j.cn112147-20231109-00300 [DOI] [PubMed] [Google Scholar]
- 5.Jonas DE, Reuland DS, Reddy SM, et al. Screening for lung cancer with low-dose computed tomography. JAMA. 2021;325(10):971-987. doi: 10.1001/jama.2021.0377 [DOI] [PubMed] [Google Scholar]
- 6.Gao W, Wen CP, Wu A, Welch HG. Association of computed tomographic screening promotion with lung cancer overdiagnosis among Asian women. JAMA Intern Med. 2022;182(3):283-290. doi: 10.1001/jamainternmed.2021.7769 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Supplemental Methods
eReferences.
Data Sharing Statement