Clinical analysis of incidental discovery of pulmonary nodules with computed tomography: a retrospective cohort study

Introduction

According to Global Cancer Statistics 2022, among malignancies, lung cancer has the highest incidence and is the leading cause of cancer-related death worldwide (1). The survival rate of lung cancer varies greatly according to stage, with the 5-year overall survival ranging from approximately 7% in stage IVB to 82% for stage IA (2). Therefore, early diagnosis and treatment are crucial for improving the prognosis of patients with lung cancer. Computed tomography (CT) is an effective means of detecting lung cancer at an early stage. The National Lung Screening Trial (NLST) reported a 20.0% reduction in lung cancer mortality with regular screening via low-dose computed tomography (LDCT) compared to chest radiography (3). Subsequently, the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) study also demonstrated the value of CT screening in reducing the lung cancer mortality rate (4). In addition, a recent multicenter prospective cohort study conducted in China found that lung cancer mortality was 31% lower in the LDCT-screened group (5), suggesting potential benefits beyond current risk-stratified approaches. CT serves as a pivotal non-invasive imaging modality for pulmonary nodule assessment and early lung cancer detection, offering rapid comprehensive diagnostic information while minimizing patient discomfort (6). However, screening for lung cancer using a CT scan is associated with certain risks such as greater radiation exposure, a higher workload for radiologists, and higher costs. Most studies on this subject have only conducted CT screening in participants at high risk according to different criteria, including smoking history, age, and other factors (7,8). However, the incidence rate of lung cancer in low-risk participants without screening was reported to be 0.3% in a large-scale study, which is not particularly low (5). Therefore, the baseline situation of CT screening in the entire population needs to be examined in a study with a relatively larger sample size.

Given the aforementioned concerns with chest CT screening, it is impractical to conduct it on a large scale for a relatively healthy population in daily clinical work. However, the coronavirus disease 2019 (COVID-19) pandemic represented a unique opportunity. During this period, all emergency patients and outpatients ready to be admitted to the wards were asked to accept chest CT scanning and viral tests to exclude COVID-19 pneumonia. Except for those individuals who accepted chest CT scanning due to related respiratory diseases and a history of malignant tumors, others could be defined as approximately healthy individuals who did not require chest CT examination in actual clinical practice. Thus, this study focused on this segment of patients, the imaging features of unexpectedly discovered lung lesions were analyzed, patients with high-risk lesions were informed of their condition, and follow-up was conducted to characterize the baseline CT screening in the entire population. This accidental screening cohort provides unique insights into true population-level nodule prevalence, malignancy rates without selection bias, and practical challenges of large-scale CT implementation. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-577/rc).

Methods

Patients

According to patients’ medical records or CT application forms, all patients underwent chest CT scans in Xuanwu Hospital, Capital Medical University from February 1, 2020 to February 1, 2022, except those who received CT scans due to respiratory system diseases or a history of malignant tumor, were included in this study. These patients were considered to be approximately healthy individuals who underwent CT scans only to exclude infection of COVID-19. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Xuanwu Hospital, Capital Medical University (approval No. KS2022057; approved on April 20, 2022). Individual consent for this retrospective analysis was waived.

Chest CT scan

Spiral CT images were obtained using a 64-detector CT row scanner (Brilliance, Philips Healthcare, Best, the Netherlands) and were reconstructed in overlapping contiguous 5-mm increments, with a pitch of 1.25 (9). A window level of –700 Hounsfield units (HU) and a window width of 1,500 HU was defined as the lung window, while the mediastinal window was defined as a window level of 40 HU and a window width of 350 HU. The maximum diameter of the nodule was measured on the lung window. At least two senior radiologists read the original images and wrote reports. Positive findings were defined as any nodule with a diameter ≥4 mm. The pulmonary nodules were further differentiated into solid nodules, part-solid nodules, and ground-glass (non-solid) nodules (GGNs). A solid nodule has a homogeneous soft-tissue attenuation, a GGN has hazy increased attenuation that does not obscure the bronchial and vascular margins, and a part-solid nodule has elements of both solid and GGNs (10).

The definition of high-risk patients and follow-up procedures

Patients with noncalcified nodules ≥6 mm identified by CT screening are defined as high risk. We informed these patients of their CT scan results over the telephone and recommended they undergo reexamination or directly receive invasive procedures to confirm the pathologic diagnosis according to the National Comprehensive Cancer Network (NCCN) lung cancer screening guidelines (10). For these patients, follow-up was conducted every 6 months via CT or immediate biopsy until a clear pathological diagnosis was made. The final follow-up date was June 30, 2024. The stage of lung cancer was decided based on the ninth edition of the TNM classification of lung cancer (2).

Statistical analysis

SPSS 26 (IBM Corp., Armonk, NY, USA) was used for all statistical analyses. Categorical variables are reported as numbers and percentages. Their differences were analyzed with the Pearson χ² test or the Fisher exact test. Statistical significance was defined as a two-sided P value <0.01 to account for increased type I error risk due to the large sample size.

Results

Between February 1, 2020 and February 1, 2022, a total of 211,759 patients underwent chest CT scans in our institution. Of these, 91,136 were excluded from the study because they received thoracic CT scans due to respiratory diseases or a history of malignant tumors. The remaining 120,623 patients underwent CT only to rule out pneumonia and were therefore included in the study. Among these patients, 63,775 (52.9%) were identified with positive results. Subsequently, the images of all patients with positive results were reviewed by senior thoracic surgeons, and a total of 6,117 patients were identified as having high-risk lesions. Subsequently, all high-risk patients were informed of their CT screening results by telephone and were advised to undergo follow-up examinations or invasive procedures to confirm the pathological diagnosis following the NCCN lung cancer screening guidelines. A total of 635 patients were excluded from follow-up due to death, loss of contact, advanced age, or other serious illnesses, while follow-up data were successfully obtained from the remaining 5,482 patients. After a median follow-up of 30 months, 343 patients received invasive procedures, and a total of 310 lung cancer and 4 metastatic carcinoma cases were ultimately confirmed by pathology. The flowchart for the whole study is shown in Figure 1.

Figure 1 Flowchart of the study. CT, computed tomography.

Among the entire screening population, the proportion of individuals with positive results was 52.9% (63,775/120,623). The incidence of positive findings did not differ significantly between genders but increased with age (Table 1, Figure 2). Subsequently, based on the definition of high-risk patients as those with noncalcified nodules ≥6 mm, a total of 6,117 cases were placed in the high-risk group, while the other 57,658 participants were placed in the low-risk group. The specific characteristics of all patients with positive results and the comparison between the high- and low-risk groups are shown in Table 2. We found that compared with the low-risk cases, high-risk cases were older, with a higher proportion of females, a higher proportion of family history of cancer, and a higher proportion of GGNs or partially solid nodules as compared to solid nodules. However, there was no significant difference in the history of heavy smoking between the two groups of patients.

Figure 2 The detection rate of positive results by age and gender.

According to the NCCN guidelines, we recommended regular follow-up or invasive diagnosis for the 6,117 patients classified as high-risk. A total of 635 cases declined to participate, while the other 5,482 agreed to participate. During a median follow-up period of 30 months, 343 patients received invasive diagnostic procedures. A total of 314 patients were diagnosed with malignant lesions, of which 310 were primary lung cancer and 4 were metastatic lesions. We further analyzed the characteristics of patients confirmed with primary lung cancer (Table 3). We discovered that a higher proportion of lung cancer patients were female, with the vast majority having early-stage lung adenocarcinoma.

Discussion

Chest CT screening has been confirmed to significantly reduce lung-cancer-related death in high-risk patients (3,5). However, the criteria for high-risk populations varies across different guidelines, such as the NCCN, International Early Lung Cancer Action Program (I-ELCAP), the United States Preventive Services Task Force (USPSTF), and China Guideline for the Screening and Early Detection of Lung Cancer (CGSL) (10-12). Therefore, large-scale CT screening studies targeting the entire population would be highly valuable, and the special circumstances surrounding the COVID-19 pandemic provided a unique opportunity. Our study provides insight drawn from opportunistic CT screening for approximately healthy individuals and may help refine inclusion criteria in the future.

Given the development of lung cancer in the population with negative initial screening results during follow-up and the refusal of cases with high-risk results, the incidence of lung cancer in our cohort was similar to the results of previous large-scale studies conducted in China (5,13). Our results indicated that among patients confirmed with lung cancer, the proportion of females was higher, and the vast majority of lesions on imaging were GGNs, which is consistent with the results of previous studies (14,15). This association may stem from exposure to second-hand smoke (SHS) and polycyclic aromatic hydrocarbons, as noted in prior research (16). Moreover, other research has suggested there to be higher malignancy in GGNs as compared to solid nodules, despite their slow progression (17). Therefore, female individuals in whom GGNs are detected may require greater clinical attention.

Within the cohort, adenocarcinoma was the predominant histological type of lung cancer, accounting for 93.9% (291 out of 310) of all patients. This result aligns with the findings of other clinical trials, with the percentage ranging from 60.6% to 91.3% (4,5,18). Correspondingly, the detection rate (5.3%) of squamous-cell carcinomas (SCCs), accounting for the majority of centrally located tumors, was lower than expected in our study or previous researches (3-5,18,19). This may be related to the fact that patients with central SCCs often present symptoms such as cough, sputum, or hemoptysis and thus seek advice from the respiratory department. In our study, this tendency is more pronounced precisely because we excluded this group of patients.

The rate of early-stage lung cancer (stage I or earlier) in our study was relatively high at 81.3%, surpassing the rate reported in the NLST (63%), NELSON trial (58.6%), and Li et al.’s study (62.7%) (3-5); meanwhile, it was similar to the results of Tang et al.’s and Zhang et al.’s studies (18,20). Moreover, the rate of stage IV lung cancer was very low (1.3%) in our study. Considering the high diagnostic rate of stage IV lung cancer in the non-screening group reported previously (2), we have reason to believe that CT screening provides survival benefits to patients, even if long-term follow-up is not conducted after a pathological diagnosis.

In our study, senior thoracic surgeons reviewed all patients’ imaging and simply defined all noncalcified nodules ≥6 mm as high risk, but this was not sufficiently accurate and extremely time-consuming. More and more studies confirmed that novel techniques such as artificial intelligence (AI) provide value for applications in lung nodules and lung cancer in fields such as automated detection, noninvasive characterization of nodules, lung cancer risk assessment, and prediction of prognosis (6,21-24). Therefore, we believe the application of these techniques may hold substantial potential in fostering the implementation of lung cancer screening.

Our findings demonstrate that CT screening detects early-stage lung cancer in the Chinese cohort, but routine population-wide screening remains impractical due to high false-positive rates (47.1% benign nodules) and resource constraints. Instead, we advocate for prioritizing high-risk groups while integrating incidental CT scans. The high proportion of adenocarcinoma (93.9%) and early-stage cancers supports refining risk models for Asian populations, incorporating environmental and familial factors. Future efforts should optimize follow-up protocols for 4–6 mm nodules. These data reinforce targeted screening over universal approaches, balancing early diagnosis with overdiagnosis risks.

Several limitations should be acknowledged in our study. First, all patients enrolled in the study were patients seeking medical treatment and were not volunteers. Therefore, they cannot be fully recognized as healthy individuals. Although we excluded patients with respiratory symptoms or a history of malignant tumors, subclinical or resolved COVID-19-related lung changes might have contributed to false-positive nodule identifications in some asymptomatic individuals. This could lead to a modest overestimation of pulmonary nodule prevalence compared to non-pandemic screening cohorts. However, our stringent adherence to Lung Imaging Reporting and Data System (Lung-RADS) criteria for malignancy-predictive features and longitudinal growth assessment likely mitigated this bias for high-risk lesions. Moreover, the proportion of individuals who refused to participate in our follow-up was relatively high due to factors such as advanced age, presence of severe primary diseases, and lack of trust. Moreover, the use of telephone follow-up might have introduced bias. In addition, we did not follow up on cases with negative screening results and low-risk lesions, so we could not determine the incidence of lung cancer in this group of patients during the same follow-up period. Furthermore, we did not continue to follow up with patients confirmed with lung cancer to obtain the long-term prognosis.

Conclusions

Based on the chest CT screening data of a relatively large asymptomatic population, there is a high probability of the accidental discovery of pulmonary nodules. A considerable number of lesions were suspected of malignancy; therefore, regular follow-up and close monitoring were necessary.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-577/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-577/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-577/prf

Funding: This study was supported by the Capital Health Development Research Special Project (No. 2022-2-2013).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-577/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Xuanwu Hospital, Capital Medical University (approval No. KS2022057; approved on April 20, 2022). Individual consent for this retrospective analysis was waived. The study was registered in the National Health Security Information Platform/Medical Research Registration and Filing System (www.medicalresearch.org.cn) (No. MR-11-22-005037).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Rami-Porta R, Nishimura KK, Giroux DJ, et al. The International Association for the Study of Lung Cancer Lung Cancer Staging Project: Proposals for Revision of the TNM Stage Groups in the Forthcoming (Ninth) Edition of the TNM Classification for Lung Cancer. J Thorac Oncol 2024;19:1007-27. [Crossref] [PubMed]
3. National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365:395-409. [Crossref] [PubMed]
4. de Koning HJ, van der Aalst CM, de Jong PA, et al. Reduced Lung-Cancer Mortality with Volume CT Screening in a Randomized Trial. N Engl J Med 2020;382:503-13. [Crossref] [PubMed]
5. Li N, Tan F, Chen W, et al. One-off low-dose CT for lung cancer screening in China: a multicentre, population-based, prospective cohort study. Lancet Respir Med 2022;10:378-91. [Crossref] [PubMed]
6. Jiang Q, Sun H, Chen Q, et al. High-resolution computed tomography with 1,024-matrix for artificial intelligence-based computer-aided diagnosis in the evaluation of pulmonary nodules. J Thorac Dis 2025;17:289-98. [Crossref] [PubMed]
7. Ji G, Bao T, Li Z, et al. Current lung cancer screening guidelines may miss high-risk population: a real-world study. BMC Cancer 2021;21:50. [Crossref] [PubMed]
8. Zhang Z, Gao Y, Liu S, et al. Initial low-dose computed tomography screening results and summary of participant characteristics: based on the latest Chinese guideline. Front Oncol 2023;13:1085434. [Crossref] [PubMed]
9. Yang W, Qian F, Teng J, et al. Community-based lung cancer screening with low-dose CT in China: Results of the baseline screening. Lung Cancer 2018;117:20-6. [Crossref] [PubMed]
10. Wood DE, Kazerooni EA, Aberle D, et al. NCCN Guidelines® Insights: Lung Cancer Screening, Version 1.2022. J Natl Compr Canc Netw 2022;20:754-64. [Crossref] [PubMed]
11. US Preventive Services Task Force. Screening for Lung Cancer: US Preventive Services Task Force Recommendation Statement. JAMA 2021;325:962-70. [Crossref] [PubMed]
12. International Early Lung Cancer Action Program Investigators. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 2006;355:1763-71. [Crossref] [PubMed]
13. van Klaveren RJ, Oudkerk M, Prokop M, et al. Management of lung nodules detected by volume CT scanning. N Engl J Med 2009;361:2221-9. [Crossref] [PubMed]
14. Li X, Ren F, Wang S, et al. The Epidemiology of Ground Glass Opacity Lung Adenocarcinoma: A Network-Based Cumulative Meta-Analysis. Front Oncol 2020;10:1059. [Crossref] [PubMed]
15. Zhang Y, Jheon S, Li H, et al. Results of low-dose computed tomography as a regular health examination among Chinese hospital employees. J Thorac Cardiovasc Surg 2020;160:824-831.e4. [Crossref] [PubMed]
16. Subramanian J, Govindan R. Lung cancer in never smokers: a review. J Clin Oncol 2007;25:561-70. [Crossref] [PubMed]
17. Migliore M, Fornito M, Palazzolo M, et al. Ground glass opacities management in the lung cancer screening era. Ann Transl Med 2018;6:90. [Crossref] [PubMed]
18. Zhang Y, Qian F, Teng J, et al. China lung cancer screening (CLUS) version 2.0 with new techniques implemented: Artificial intelligence, circulating molecular biomarkers and autofluorescence bronchoscopy. Lung Cancer 2023;181:107262. [Crossref] [PubMed]
19. Paci E, Puliti D, Lopes Pegna A, et al. Mortality, survival and incidence rates in the ITALUNG randomised lung cancer screening trial. Thorax 2017;72:825-31. [Crossref] [PubMed]
20. Tang W, Liu L, Huang Y, et al. Opportunistic lung cancer screening with low-dose computed tomography in National Cancer Center of China: The first 14 years' experience. Cancer Med 2024;13:e6914. [Crossref] [PubMed]
21. de Margerie-Mellon C, Chassagnon G. Artificial intelligence: A critical review of applications for lung nodule and lung cancer. Diagn Interv Imaging 2023;104:11-7. [Crossref] [PubMed]
22. Yang S, Lim SH, Hong JH, et al. Deep learning-based lung cancer risk assessment using chest computed tomography images without pulmonary nodules ≥8 mm. Transl Lung Cancer Res 2025;14:150-62. [Crossref] [PubMed]
23. Zhang H, Tu W, Zhang Z, et al. Effect of respiratory phase on three-dimensional quantitative parameters of pulmonary subsolid nodules in low-dose computed tomography screening for lung cancer. J Thorac Dis 2025;17:1580-92. [Crossref] [PubMed]
24. Guo C, Zhang X, Shen S, et al. Differentiation of inflammatory pseudotumors and malignant pulmonary nodules using the time-to-peak in first-pass dual-input volume computed tomography-perfusion. Quant Imaging Med Surg 2025;15:2754-65. [Crossref] [PubMed]

(English Language Editor: J. Gray)