Outcomes of lung cancer screening by low-dose computed tomography in the health screening program

Abstract

Background

Low-dose computed tomography (LDCT) screening reduces lung cancer mortality among high-risk smokers, but data from Southeast Asia remain limited. This study evaluated the detection yield and risk factors of lung cancer among adults undergoing LDCT health screening in Thailand.

Methods

We retrospectively reviewed 4,478 adults aged 18–85 years who underwent LDCT between January 2019 and December 2023 at a tertiary private hospital in Bangkok. Scans were performed using a Philips IQon Spectral CT (256-slice, 120 kV, 30 mA, 1 mm slice thickness, radiation dose <1.5 mSv) and interpreted according to Lung Imaging Reporting and Data System (Lung-RADS) version 1.1. Lung-RADS 3–4X were confirmed histopathologically. Multivariable logistic regression identified predictors of lung cancer.

Results

Among all participants, 1,980 (44.2%) were female and 84.0% were never-smokers, with a mean age of 56.9 years [standard deviation (SD) 12.5]. LDCT positivity was 55.8%, and 53 cancers (1.2%) were detected. Most were stage 0–IB (69.8%), with adenocarcinoma as the predominant histology (85%) and solid morphology (77.4%). Incidental findings occurred in 65.2%, most commonly coronary artery calcification (CAC) (39.8%). Independent predictors included age ≥55 years [adjusted odds ratio (OR) 4.82, 95% confidence interval (CI): 1.68–13.86], smoking history (P<0.001), CAC (adjusted OR 23.63, 95% CI: 2.09–266.98), and extrapulmonary malignancy (adjusted OR 173.54, 95% CI: 42.73–704.77).

Conclusions

LDCT health screening in an unselected Thai population detected a meaningful burden of early-stage lung cancer. The findings support age-based screening beginning at 55 years, highlighting the potential role of chronic inflammation and environmental injury, such as PM_2.5 exposure and prior tuberculosis.

Highlight box.

Key findings

• In this 5-year retrospective study of 4,478 adults undergoing low-dose computed tomography (LDCT) as part of a health-screening program in Thailand, lung cancer was detected in 1.2% of participants, despite the majority (84%) being never-smokers.

• Most cancers (69.8%) were detected at stage 0–IB, and adenocarcinoma was the predominant histology (85%).

• Independent predictors of lung cancer included age ≥55 years, smoking history, coronary artery calcification, and extrapulmonary malignancy.

What is known and what is new?

• Previous large LDCT trials, such as the National Lung Screening Trial and the Dutch-Belgian Randomized Lung Cancer Screening Trial demonstrated mortality reduction in high-risk smokers but included few Asian participants and almost no never-smokers.

• Evidence from Southeast Asia remains limited. This study adds real-world data from a predominantly never-smoking Thai population, showing that LDCT can detect a meaningful proportion of early-stage, resectable lung cancers even outside traditional high-risk groups.

What is the implication, and what should change now?

• These findings support expanding LDCT screening consideration to Asian populations beyond smoking-based criteria, particularly for adults aged ≥55 years.

• Incorporating cardiovascular and chronic inflammatory markers such as coronary calcification may improve individualized risk stratification for future screening protocols.

Introduction

Lung cancer remains the leading cause of cancer-related mortality worldwide. Although cigarette smoking accounts for most cases, the burden of lung cancer among never-smokers is disproportionately high in Asia (1).

Low-dose computed tomography (LDCT) has proven effective in reducing lung cancer mortality among high-risk smokers in large randomized trials such as the National Lung Screening Trial (NLST) and the Dutch-Belgian Randomized Lung Cancer Screening Trial (NELSON) (1,2).

The Taiwan Lung Cancer Screening in Never-Smoker Trial (TALENT) also found similar finding through a regional, multicenter cohort conducted across 17 tertiary hospitals, enrolling adults aged 55–75 years with low or no smoking history and additional risk factors such as a family history of lung cancer, prior tuberculosis (TB), or chronic exposure to cooking fumes. TALENT reported a 2.6% of lung cancer prevalence with LDCT and identified family history and female sex as major predictors of invasive lung cancer. In Taiwan, nearly 60% of never-smokers are diagnosed at stage IV, underscoring the importance of screening in this never-smoker population (3).

Recent data from China’s Guangzhou Lung-Care Project extended this paradigm, showing that non-risk-based LDCT screening including younger and non-smoking individuals yielded comparable prevalence of lung cancer (1.7%) and a high proportion of stage I disease, suggesting that traditional risk-based models may fail to capture a substantial proportion of Asian cases (4). Building on this evidence, the Asian Expert Consensus on LDCT Screening [2023] now emphasizes region-specific eligibility criteria, integrating environmental exposure, TB history, and family history of lung cancer as potential risk factors (5).

Evidence from East Asia supports LDCT screening beyond traditional smoking-based criteria, yet its relevance to Southeast Asia is uncertain (5). Thailand, with its high TB prevalence (6) and rapid expansion of private LDCT screening, represents an important setting to examine real-world performance.

Evidence on the use of LDCT screening in Southeast Asian populations remains limited, particularly among individuals with low smoking exposure. To address this gap, this study aimed to evaluate the diagnostic yield and spectrum of findings from LDCT screening in a general health check-up population in Thailand, regardless of smoking status, with the goal of informing population-appropriate screening strategies and refining eligibility criteria for lung cancer screening in Southeast Asian settings. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2450/rc).

Methods

Study design and setting

We conducted a retrospective cohort study of adults who underwent LDCT as part of a self-referred health screening program, offered as a commercial check-up package for asymptomatic individuals at a private tertiary-care hospital in Bangkok, Thailand, between January 1, 2019, and December 31, 2023. In this context, health screening refers to routine preventive imaging performed at the individual’s discretion, without physician referral or clinical symptoms. LDCT examinations were identified through the hospital’s radiology information system (RIS) and picture archiving and communication system (PACS). Clinical and demographic data were obtained from the electronic medical record. This study was approved by the Bangkok Hospital Headquater (BHQ) Institutional Review Board (IRB No. 2024-12-53). The requirement for informed consent was waived because the study used de-identified retrospective data. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Participants and procedures

Eligible participants were adults aged 18–85 years, irrespective of smoking history, who underwent LDCT for health screening by self-referred covered by private insurance or self paid. Smoking history was self-reported and classified as current smoker (smoked within the past 30 days), ex-smoker (previously smoked but not within the past 30 days), or never (had never smoked).

We excluded individuals with a prior diagnosis of lung cancer, coronavirus disease 2019 (COVID-19) pneumonia, or age <18 or >85 years. All LDCT examinations were performed using a standardized protocol (detailed below).

Computed tomography (CT) acquisition and image interpretation protocol

All LDCT examinations were performed using a Philips IQon Spectral CT scanner (Philips Healthcare, The Netherlands) equipped with 256-slice capability and a 128-detector configuration. Scans were obtained in the helical mode with a tube voltage of 120 kV and tube current of 30 mA. Participants were positioned supine and scanned in the superior-to-inferior direction during a full-inspiration breath-hold without intravenous contrast administration. The scan range covered the entire thorax.

Images were reconstructed with a slice thickness of 1 mm using spectral CT acquisition. Reconstruction parameters included a lung window width of 1,600 HU and window level −500 to 1,500 HU, and a soft-tissue window width of 360 HU and window level 40 to 400 HU. Multiplanar reformats were generated in axial, coronal, and sagittal planes. The effective radiation dose for each LDCT examination was less than 1.5 mSv.

All LDCT images and reports were retrospectively reviewed by one board-certified radiologist to reduce inter-observer variability. Image interpretation followed the Lung Imaging Reporting and Data System (Lung-RADS) version 1.1 for nodule categorization and follow-up recommendations.

All LDCT findings were categorized according to the Lung-RADS version 1.1 to standardize the interpretation and reporting of pulmonary nodules. Lung-RADS classifies findings based on nodule size, composition, and morphology: Category 1 indicates no nodules or benign calcified nodules; Category 2, solid or part-solid nodules smaller than 6 mm or ground-glass nodules (GGNs) smaller than 20 mm; Category 3, nodules measuring 6–8 mm or GGNs ≥20 mm; Category 4A, solid nodules 8–15 mm or part-solid nodules with solid components 6–8 mm; Category 4B, solid nodules ≥15 mm or part-solid nodules with solid components ≥8 mm; and Category 4X, Category 3 or 4 nodules with additional suspicious features (7).

For each participant, Lung-RADS categories were assigned at the time of reporting. A positive LDCT was defined as any scan categorized as Lung-RADS 2, 3, 4A, 4B, or 4X. Cases with suspected lung cancer or mimicking lesions were further evaluated in consensus with two pulmonologists. Participants with positive findings were followed up for 3 years, and diagnoses of lung cancer were confirmed by histopathology from image-guided biopsy or surgical resection.

Outcomes

The primary outcome was the number and proportion of participants diagnosed with lung cancer during the study period. Secondary outcomes included stage at diagnosis [based on the International Association for the Study of Lung Cancer (IASLC) 9th edition tumor-node-metastasis (TNM) classification (8)], the number of intrathoracic and extrathoracic malignancies, and the frequency of benign findings such as granulomas or non-neoplastic nodules.

Statistical analysis and sample size

Continuous variables were summarized as means with standard deviations (SDs) for normally distributed data and as medians with interquartile ranges (IQRs) for non-normally distributed data. Normality was assessed using the Shapiro-Wilk test. When the distribution was not normal, nonparametric tests (Mann-Whitney U test) were applied instead of the two-sample t-test. Categorical variables were summarized as counts and percentages and compared using the χ² test or Fisher’s exact test as appropriate. Detection rates were expressed as percentages with 95% confidence intervals (CIs) calculated using the Wald method (9). The rate ratio (RR) and its 95% CI were estimated using unconditional maximum likelihood. Missing data were minimal and had no meaningful impact on the analysis.

The sample size was calculated based on an expected lung cancer detection rate of 1.1%, derived from the baseline low-dose CT detection rate reported in the NLST (1), using a two-sided α of 0.05 and 80% power (β=0.20), resulting in a required sample size of 4,478 participants.

Results

Between January 1, 2019, and December 31, 2023, a total of 4,478 participants underwent LDCT. The baseline characteristics are summarized (Table 1). Of these participants, 1,980 (44.2%) were female, and the mean (SD) age was 56.9 (12.5) years (range: 18–85 years). The majority [3,763 (84.0%)] were never-smokers, and the cohort was predominantly Thai (66.0%), followed by Caucasian (13.7%), other Asian (12.8%), Hispanic (6.9%), and African (0.6%). A total of 710 participants (15.9%) reported a history of chronic lung disease, including pulmonary TB, chronic obstructive pulmonary disease (COPD), asthma, and bronchiectasis.

Table 1. Baseline characteristics of all participants in the LDCT screening cohort.

Characteristics	All participants (n=4,478)
Sex
Female	1,980 (44.2)
Male	2,498 (55.8)
Group age (years)
<40	457 (10.2)
41–50	871 (19.5)
51–60	1,381 (30.8)
61–70	1,090 (24.3)
>70	679 (15.2)
Age (years)
Mean (SD)	56.9 (12.5)
Range	18–85
Group nationality
Thai	2,956 (66.0)
Caucasian	614 (13.7)
Others Asian	573 (12.8)
Hispanic	307 (6.9)
African	28 (0.6)
Smoking
Never	3,763 (84.0)
Currently	516 (11.5)
Quit	199 (4.4)
History of chronic lung disease
Asthma	323 (7.2)
COPD/emphysema	150 (3.3)
TB	80 (1.8)
Bronchiectasis	283 (6.3)

Data are presented as n (%) unless otherwise indicated. COPD, chronic obstructive pulmonary disease; LDCT, low-dose computed tomography; SD, standard deviation; TB, tuberculosis.

Among all 4,478 baseline LDCT scans, 2,500 (55.8%) were considered positive findings (Table 2). LDCT findings were distributed as follows: Lung-RADS 1—1,978 (44.2%), Lung-RADS 2—2,155 (48.1%), Lung-RADS 3—206 (4.6%), Lung-RADS 4A—60 (1.3%), Lung-RADS 4B—52 (1.2%), and Lung-RADS 4X—27 (0.6%).

Table 2. LDCT findings, cancer characteristics, and pathologic features of participants with and without lung cancer.

Characteristics	Lung cancer (n=53)	Non-lung cancer (n=4,425)
LDCT finding
Pulmonary nodule, LUNG-RADS 1	–	1,978 (44.7)
Pulmonary nodule, LUNG-RADS 2	1 (1.9)	2,154 (48.7)
Pulmonary nodule, LUNG-RADS 3	1 (1.9)	205 (4.6)
Pulmonary nodule, LUNG-RADS 4A	7 (13.2)	53 (1.2)
Pulmonary nodule, LUNG-RADS 4B	21 (39.6)	31 (0.7)
Pulmonary nodule, LUNG-RADS 4X	23 (43.4)	4 (0.1)
Positive CT finding (LUNG-RADS 2–4)	53 (100.0)	2,447 (55.3)
Negative CT finding (LUNG-RADS 1)	–	1,978 (44.7)
Staging
Stage 0	1 (1.9)
Stage IA1	4 (7.5)
Stage IA2	20 (37.7)
Stage IA3	9 (17.0)
Stage IB	3 (5.7)
Stage IIA	1 (1.9)
Stage IIB	2 (3.8)
Stage IIIA	0 (0.0)
Stage IIIB	3 (5.7)
Stage IIIC	0 (0.0)
Stage IVA	7 (13.2)
Stage IVB	3 (5.7)
Imaging
Solid	41 (77.4)
Part solid	10 (18.9)
Ground glass	2 (3.8)
Pathology report
Adenocarcinoma in situ	1 (1.9)
Minimally invasive adenocarcinoma	13 (24.6)
Invasive adenocarcinoma	31 (58.5)
Squamous cell carcinoma	4 (7.5)
MALT lymphoma	1 (1.9)
Small cell carcinoma	1 (1.9)
Neuroendocrine	1 (1.9)
Large cell carcinoma	1 (1.9)
Tumor size, mm
<6	0 (0.0)
6–8	1 (1.9)
>8, ≤15	17 (32.1)
>15, ≤30	24 (45.3)
>30	11 (20.8)

Data are presented as n (%). CT, computed tomography; LDCT, low-dose computed tomography; LUNG-RADS, Lung Imaging Reporting and Data System; MALT, mucosa-associated lymphoid tissue.

Overall, 53 participants (1.2%) were diagnosed with lung cancer, confirmed by bronchoscopy, transthoracic biopsy, or surgical pathology. Of these, 40 (75.5%) were detected on the initial LDCT, while 8 cases demonstrated interval growth over 7 months to 5 years of follow-up. Three nodules progressed from ground-glass to part-solid or solid appearance within 2–5 years, and two remained stable for up to 2 years. All lung cancers occurred in participants aged ≥40 years. Early-stage disease predominated: 37 cases (69.8%) were stage 0–IB. By imaging morphology, 41 (77.4%) were solid, 10 (18.9%) were part-solid, and 2 (3.8%) were GGNs, with 41.5% of solid nodules showing spiculated margins. Histopathologic diagnoses included adenocarcinoma in situ [1 (1.9%)], minimally invasive adenocarcinoma [13 (24.6%)], invasive adenocarcinoma [31 (58.5%)], squamous cell carcinoma [4 (7.5%)], and other malignancies such as large cell carcinoma, small cell carcinoma, and mucosa-associated lymphoid tissue (MALT) lymphoma [4 (7.5%) in total]. One patient (1.9%) had synchronous multiple primary lung cancers (sMPLC). When stratified by smoking history, lung cancer detection rates differed across groups. Among never-smokers (n=3,763), lung cancer was detected in 31 participants (0.82%). In current smokers (n=516), 14 participants (2.71%) were diagnosed with lung cancer. The highest detection rate was observed among former smokers (quitters; n=199), with lung cancer identified in 8 participants (4.02%).

Additionally, 8 participants (0.2%) had pulmonary metastases from non-lung primary cancers, including colon [3], rectal [1], breast [1], transitional cell carcinoma of the renal pelvis [1], leiomyosarcoma [1], and malignant meningioma [1]. Two benign pulmonary tumors (carcinoid tumor and squamous papilloma) were also identified.

Extra-pulmonary malignancies were incidentally detected in 26 participants (0.58%), including thyroid [8], renal cell carcinoma [7], breast (4; one bilateral), hepatocellular carcinoma [2], intra-abdominal lymphoma [2], posterior mediastinal lymphoma [1], transitional cell carcinoma of the renal pelvis [1], and dermatofibrosarcoma protuberans of the back [1].

Active or prior TB was detected in 51 participants (1.1%)—6 active (0.14%) and 45 old (1.02%)—and other active lung infections were found in 88 (1.99%). Seven cases (0.16%) initially suspected of lung cancer were later confirmed as infection-related lesions after biopsy or follow-up imaging. To determine the most appropriate age threshold for risk stratification, we explored several cut-off points (40, 45, 50, 55, 60, 65, and 70 years) in univariate logistic regression analyses (Table S1). As shown in Table 3, the adjusted odds ratio (OR) increased progressively with age, reaching its highest and most statistically robust value at 55 years (adjusted OR 4.82, 95% CI: 1.68–13.86; P=0.004). Based on this pattern, 55 years old was selected as the optimal cut-off point for defining the higher-risk group in subsequent analyses.

Table 3. Risk factors for lung cancer and adjusted ORs after multivariable analysis.

Characteristics	Non-lung cancer (n=4,425)	Lung cancer (n=53)	Crude OR			Adjusted OR
			OR (95% CI)	P value		OR (95% CI)	P value
Sex				0.40			0.19
Male	2,465 (55.7)	33 (62.3)	1.31 (0.75–2.29)			0.58 (0.26–1.30)
Female	1,960 (44.3)	20 (37.7)	Reference			Reference
Age, years				<0.001***			0.004**
≥55	2,375 (53.7)	47 (88.7)	6.76 (2.89–15.85)			4.82 (1.68–13.86)
<55	2,050 (46.3)	6 (11.3)	Reference			Reference
Ethnicity							0.29 (Chi-squared)
Thai	2,923 (66.1)	33 (62.3)	–
Caucasian	608 (13.7)	6 (11.3)	–
Others Asian	567 (12.8)	6 (11.3)	–
Hispanic	300 (6.8)	7 (13.2)	–
African	27 (0.6)	1 (1.9)	–
Thai or others				0.56			0.65
Thai	2,923 (66.1)	33 (62.3)	1.18 (0.67–2.06)			0.86 (0.48–1.55)
Others	1,502 (33.9)	20 (37.7)	Reference			Reference
Smoking history							<0.001*** (Chi-squared)
Current smoker	502 (11.3)	14 (26.4)	–
Never	3,732 (84.3)	31 (58.5)	–
Ex-smoker	191 (4.3)	8 (15.1)	–
Asthma				0.10			0.93
Yes	316 (7.1)	7 (13.2)	1.98 (0.89–4.41)			1.05 (0.31–3.61)
No	4,109 (92.9)	46 (86.8)	Reference			Reference
Bronchiectasis				0.15			0.33
Yes	277 (6.3)	6 (11.3)	1.91 (0.81–4.51)			0.46 (0.09–2.22)
No	4,148 (93.7)	47 (88.7)	Reference			Reference
Emphysema				<0.001***			0.06
Yes	142 (3.2)	8 (15.1)	5.36 (2.48–11.59)			3.40 (0.94–12.30)
No	4,283 (96.8)	45 (84.9)	Reference			Reference
Pulmonary tuberculosis				0.62			0.005**
Yes	79 (1.8)	1 (1.9)	1.058 (0.14–7.75)			0.002 (0.00–0.17)
No	4,346 (98.2)	52 (98.1)	Reference			Reference
Fibro calcified lesions compatible with old pulmonary tuberculosis				0.001**			<0.001***
Yes	3 (0.1)	2 (3.8)	57.80 (9.46–353.33)			153.45 (18.58–1,267.55)
No	4,422 (99.9)	51 (96.2)	Reference			Reference
COPD				<0.001***			0.29
Yes	23 (0.5)	5 (9.4)	19.94 (7.28–54.63)			0.28 (0.03–2.97)
No	4,402 (99.5)	48 (90.6)	Reference			Reference
Coronary artery calcification				<0.001***			0.01*
Yes	5 (0.1)	12 (22.6)	258.73 (87.20–767.73)			23.63 (2.09–266.98)
No	4,420 (99.9)	41 (77.4)	Reference			Reference
Ischemic heart disease				<0.001***			0.48
Yes	46 (1.0)	9 (17.0)	19.47 (8.98–42.21)			2.00 (0.30–13.52)
No	4,379 (99.0)	44 (83.0)	Reference			Reference
Arrhythmia				<0.001***			0.76
Yes	42 (0.9)	6 (11.3)	13.32 (5.40–32.85)			0.60 (0.02–17.08)
No	4,383 (99.1)	47 (88.7)	Reference			Reference
Heart failure				0.35			0.66
Yes	35 (0.8)	1 (1.9)	2.41 (0.32–17.94)			0.42 (0.01–20.00)
No	4,390 (99.2)	52 (98.1)	Reference			Reference
Previous stroke				<0.001***			0.24
Yes	22 (0.5)	5 (9.43)	20.85 (7.58–57.34)			3.42 (0.44–26.73)
No	4,403 (99.5)	48 (90.6)	Reference			Reference
Parkinsonism				>0.99			0.99
Yes	4 (0.1)	0 (0.0)	9.18 (0.49–172.69)			0.00 (0.00–∞)
No	4,421 (99.9)	53 (100.0)	Reference			Reference
Hypothyroidism				0.21			0.99
Yes	126 (2.8)	0 (0)	0.318 (0.02–5.17)			0.00 (0.00–∞)
No	4,299 (97.2)	53 (100)	Reference			Reference
Hyperthyroidism				0.10			0.009**
Yes	43 (1.0)	2 (3.8)	4.00 (0.94–16.94)			0.004 (0.004–0.25)
No	4,382 (99.0)	51 (96.3)	Reference			Reference
Autoimmune disease				<0.001***			0.52
Yes	13 (0.3)	3 (5.7)	20.36 (5.63–73.68)			2.22 (0.20–24.22)
No	4,412 (99.7)	50 (94.3)	Reference			Reference
Liver cirrhosis				>0.99			0.99
Yes	5 (0.1)	0 (0.0)	7.51 (0.41–137.55)			0.00 (0.00–∞)
No	4,420 (99.9)	53 (100.0)	Reference			Reference
Chronic hepatitis				>0.99			0.99
Yes	18 (0.4)	0 (0.0)	2.23 (0.13–37.43)			0.00 (0.00–∞)
No	4,407 (99.6)	53 (100.0)	Reference			Reference
Peptic ulcer				0.002**			0.12
Yes	4 (0.1)	2 (3.8)	43.34 (7.76–241.97)			10.82 (0.56–210.47)
No	4,421 (99.9)	51 (96.2)	Reference			Reference
End stage renal disease				0.006**			0.20
Yes	59 (1.3)	4 (7.5)	6.04 (2.11–17.28)			3.09 (0.55–17.46)
No	4,366 (98.7)	49 (94.5)	Reference			Reference
Diabetes				<0.001***			0.55
Yes	466 (10.5)	15 (28.3)	3.35 (1.83–6.14)			1.37 (0.49–3.83)
No	3,959 (89.5)	38 (71.7)	Reference			Reference
Hypertension				0.002**			0.47
Yes	1,287 (29.1)	26 (49.1)	2.35 (1.37–4.04)			0.72 (0.30–1.74)
No	3,138 (70.9)	27 (50.9)	Reference			Reference
Extrapulmonary malignancy				<0.001***			<0.001***
Yes	7 (0.2)	18 (34.0)	324.59 (127.52–826.17)			173.54 (42.73–704.77)
No	4,418 (99.8)	35 (66.0)	Reference			Reference
History of extra pulmonary malignancy				<0.001***			0.29
Yes	3 (0.1)	3 (5.7)	88.44 (17.43–448.87)			0.19 (0.01–4.12)
No	4,422 (99.9)	50 (94.3)	Reference			Reference
Family history of lung cancer				<0.001***			0.98
Yes	0 (0.0)	4 (7.6)	804.64 (42.75–15,146.59)			<0.001 (0–∞)
No	4,425 (100.0)	49 (92.5)	Reference			Reference
Family history of non-lung cancer				0.002**			0.04*
Yes	4 (0.1)	2 (3.8)	43.34 (7.76–241.97)			22.14 (1.24–395.50)
No	4,421 (99.9)	51 (96.3)	Reference			Reference
Family history of heart attack				0.04*			0.93
Yes	2 (0.1)	1 (1.9)	42.53 (3.80–476.35)			2.20 (0–∞)
No	4,423 (99.9)	52 (98.1)	Reference			Reference

Data are presented as number (%). *, P<0.05; **, P<0.01; ***, P<0.001. Crude ORs were calculated using univariate logistic regression. Variables with P<0.05 in univariate analysis were included in the multivariable logistic regression model, adjusted for age, sex, race and smoking history. A P value <0.05 was considered statistically significant. CI, confidence interval; COPD, chronic obstructive pulmonary disease; OR, odds ratio.

As summarized in Table 3, multivariable analysis identified several factors independently associated with lung cancer detection on LDCT. Participants aged ≥55 years had a significantly higher risk (adjusted OR 4.82, 95% CI: 1.68–13.86; P=0.004), and smoking history remained strongly correlated with cancer occurrence (P<0.001, χ² test). Coronary artery calcification (CAC) (adjusted OR 23.63, 95% CI: 2.09–266.98; P=0.01), extrapulmonary malignancy (adjusted OR 173.54, 95% CI: 42.73–704.77; P<0.001), and a family history of non-lung cancer (adjusted OR 22.14, 95% CI: 1.24–395.50; P=0.04) were also significant predictors of lung cancer on screening. Interestingly, while pulmonary TB (adjusted OR 0.002, 95% CI: 0.00–0.17; P=0.005) and hyperthyroidism (adjusted OR 0.004, 95% CI: 0.004–0.25; P=0.009) showed statistical significance, their inverse direction of association and extremely wide CIs indicate that these findings are likely statistical rather than clinically meaningful.

Overall, incidental findings (IF) were present in 2,920 participants (65.2%). Of these, 477 (10.8%) were intrapulmonary, and 2,689 (60.8%) were extra-pulmonary. The most common IFs included CAC [1,783 (39.8%)], fatty liver [489 (10.9%)], liver cysts [422 (9.4%)], renal cysts [337 (7.5%)], renal stones [197 (4.4%)], and gallstones [111 (2.5%)].

Discussion

Although the majority of participants were never-smokers, this 5-year retrospective study of 4,478 adults who underwent LDCT as part of a health-screening program in Thailand detected lung cancer in 1.2% of participants despite most being never-smokers, lung cancer was detected in 1.2% of participants. These findings demonstrate that LDCT screening in an unselected, predominantly never-smoking Southeast Asian population can detect a meaningful burden of early-stage, resectable lung cancer while also revealing a high rate of clinically relevant IFs.

Large, randomized trials and recent guidelines have consistently used age 50–80 years as the standard eligibility range for LDCT screening, but with notable variations in entry thresholds across studies and regions. The NLST in the United States enrolled participants aged 55–74 years with a ≥30 pack-year smoking history, demonstrating a 20% reduction in lung cancer mortality with LDCT compared to chest radiography (1). The NELSON trial in Europe included men aged 50–74 years who were current or former smokers with heavy cumulative exposure and showed a 24% mortality reduction after 10 years of follow-up (10). Both trials therefore set the evidence base for the ≥55-year cut point adopted in most Western guidelines. However, the 2023 American Cancer Society (ACS) update now recommends annual LDCT screening for adults aged 50–80 years with ≥20 pack-years of smoking history (11), lowering the age threshold to 50 years to expand access to younger high-risk populations.

In contrast to these risk-based designs, our proposed ≥55-year-old cut point was empirically derived from local data, reflecting the age at which the adjusted OR for lung cancer detection on LDCT was highest in our cohort. This cut point aligns with the NLST and NELSON entry criteria but is narrower than the current ACS recommendation, supporting its applicability in populations with lower smoking prevalence but increasing age-related risk. It also underscores the need for region-specific thresholds that balance detection yield and overdiagnosis in predominantly never-smoking Asian populations.

In interpreting our results, the difference in risk between active and old TB provides important insight into the chronic inflammatory mechanisms linking infection and lung cancer (Table 3). In our cohort, participants with newly diagnosed or active TB showed almost no increased risk, whereas those with fibro-calcified lesions compatible with old, healed TB had a markedly higher adjusted OR. This pattern is consistent with findings from Hwang et al. (12), who reported that a prior history of pulmonary TB was associated with a significantly elevated risk of lung cancer independent of smoking status and other confounders, indicating that the effect persists long after the infection has resolved. Furthermore, Moon et al. (13) conducted a large population-based cohort study and found that TB survivors had a 1.7-fold higher risk of developing lung cancer compared with controls, independent of smoking or COPD status. The authors further explained that chronic post-TB inflammation, immune dysregulation, and oxidative stress within scarred lung tissue may promote epithelial proliferation and genetic instability, thereby fostering malignant transformation. These observations suggest that the carcinogenic process arises from chronic fibrosis and inflammation following previous TB infection rather than from active disease, which may explain the strong association seen in our cohort between old, healed TB and lung cancer.

Another notable finding in our study was the strong association between CAC and lung cancer detection on LDCT. This relationship may reflect shared biological mechanisms driven by chronic exposure to fine particulate matter (PM). PM particles can penetrate into systemic and pulmonary circulation, where they initiate oxidative stress and persistent low-grade inflammation (14). These processes contribute to endothelial dysfunction, lipid oxidation, and vascular smooth muscle proliferation, ultimately leading to atherosclerotic plaque formation and coronary calcification (14). Simultaneously, within the lung, PM_2.5 exposure promotes reactive oxygen species (ROS) generation, DNA strand breaks, and upregulation of proinflammatory cytokines such as IL-1β, TNF-α, and NF-κB, all of which are implicated in cellular transformation and carcinogenesis (15). Consequently, the coexistence of CAC and lung cancer in our cohort may not represent a direct causal relationship but rather a shared outcome of prolonged PM_2.5 exposure, serving as an indicator of cumulative environmental and inflammatory burden.

This study has several limitations. It was conducted at a single private hospital with self-referred participants, which may limit generalizability to other settings. Smoking history and comorbidity data were self-reported and may include minor misclassification. Selection bias should be considered when interpreting our findings, as participants were self-referred individuals who voluntarily enrolled in a health screening program and may represent a more health-conscious population than the general public. In addition, the retrospective design did not allow evaluation of long-term outcomes such as mortality reduction. Despite these limitations, the study provides valuable real-world evidence on LDCT performance in an underrepresented Southeast Asian population and supports future prospective validation. Future studies could explore additional risk factors, including second-hand smoke exposure and cooking oil fumes, to develop a more comprehensive risk profile for LDCT lung cancer screening.

Conclusions

In conclusion, LDCT screening among an unselected, predominantly never-smoking Thai population detected a meaningful proportion of early-stage, resectable lung cancers. Age ≥55 years, smoking history, CAC, and extrapulmonary malignancy were independent predictors of lung cancer detection. These findings suggest that age-based LDCT screening beginning at 55 years may be appropriate for Asian populations.

Supplementary

The article’s supplementary files as

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 approved by the Bangkok Hospital Headquater (BHQ) Institutional Review Board (IRB No. 2024-12-53). The requirement for informed consent was waived because the study used de-identified retrospective data. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2450/dss

jtd-18-02-108-dss.pdf ^{(98.7KB, pdf)}

DOI: 10.21037/jtd-2025-1-2450