Introduction

Lung cancer (LC) is the leading cause of cancer-related death both in China and worldwide [1, 2]. Although LC screening of high-risk populations with low-dose computed tomography (LDCT) can reduce mortality [3,4,5], the promotion and application of LC screening are not very effective globally [6,7,8], with only a small portion of the population benefiting from organized nonprofit screening projects [9,10,11]. In recent years, due to various reasons, such as people’s increasing awareness of their health, opportunistic screening, and infection, an increasing number of people have received chest CT scans, so a large number of incidental pulmonary nodules (PNs) have been detected [6, 12], some of which in the non-high-risk population have been verified as LC[7, 13]. With the popularization of chest CT, the promotion of anti-smoking campaigns, continuously declining smoking rates and the aging of the population, the number of people who do not meet the present definition of a high-risk population is progressively increasing, including nonsmokers, those who seldom smoke, and those who smoked a lot in the past but quit smoking many years ago [14]. Non-high-risk individuals could become the most common form of LC. Therefore, it is imperative to understand the characteristics and outcomes of non-high-risk individuals with stage T1 LC (with a maximum tumor diameter ≤ 3 cm)

The clinical characteristics and prognosis of high-risk patients with LC have been well documented [15,16,17,18]. There are currently many studies on LC in never smokers (LCINS) [19, 20]; however, there is currently no evidence documenting the clinical characteristics and prognosis of non-high-risk patients with incidental stage T1 LC. The aim of this study was to investigate the characteristics and prognosis of patients who do not meet the US preventative services task force (USPSTF) criteria but are diagnosed with incidental stage T1 LC, which is vital for accurately assessing the degree of malignancy of non-high-risk PNs; developing strategies for prospective screening, diagnosis and treatment; and improving the prognosis of LC patients.

Materials and Methods

The data and aim

This was a single-center, prospective observational cohort study. The consecutive patients who were volunteers joining the whole process management of the PNs and LC Database (China National Copyright Administration with the software copyright registration number: 2019SR0350697) in the PNs Clinic, the First Affiliated Hospital of Chongqing Medical University from 1st Jan 2019 to 31st Dec 2023, were eligible for this study. The follow-up time for all participants concluded on Jan 31st, 2024, or upon death.

The primary outcome was to compare the overall and LC-specific survival probabilities of stage T1 LC individuals in the non-high-risk and high-risk groups. The secondary outcomes were comparisons of the baseline clinical characteristics, including demographic variables, histological types and TNM staging, between cohorts.

Cohort and design

All patients pathologically diagnosed with LC via surgery, percutaneous pulmonary puncture biopsy, bronchoscopy, or examination of pleural effusion were included in this study. The detailed inclusion criteria were patients who met the following conditions: (1) aged ≥ 18 years at enrollment in the database, (2) had a maximum tumor diameter ≤ 3 cm (as determined by lung window CT imaging), and (3) were pathologically diagnosed with LC. Patients with a mass (maximum tumor diameter > 3 cm), benign PNs, metastatic pulmonary carcinoma, or other extrathoracic tumors or who lacked TNM staging were excluded. The study received approval from the Ethics Committee of our institution (Approval Number: 2019–083). Informed consent was obtained from all participants, and confidentiality was rigorously maintained.

According to the 2021 USPSTF recommendations [15], individuals aged 50–80 years with a history of smoking for 20 pack-years were classified into the high-risk (observation) group, while others were categorized into the non-high-risk (control) group. The patients were staged according to the 8th edition TNM staging guidelines of the International Association for the Study of Lung Cancer (IASLC) [20].

Subgroup analysis

Subgroup analyses were planned for the following variables: (1) lung tumors with a maximum diameter and radiological consolidation of the maximum tumor diameter (consolidation tumor ratio, CTR); CTR < 1 was defined as the subsolid LC subgroup, while CTR = 1 was defined as the pure-solid LC subgroup; [21] (2) the pure-solid stage T1 LC surgery subgroup, where CTR = 1; and (3) the pure-solid stage Ia lung adenocarcinoma (LUAD) surgery subgroup, where CTR = 1 and stage Ia. The clinical characteristics and prognosis of non-high-risk individuals were compared with those of high-risk individuals in each of the above four subgroups.

Demographic and clinical variables

Clinical features, including indications for imaging, age, sex, smoking status, medical history (chronic obstructive pulmonary disease, COPD; tuberculosis, TB; extrathoracic tumor, other systems diseases), environmental exposure, family history of cancer, lung function, diagnostic method and timing, surgical complications, histological type, and tumor TNM staging, were extracted from our database (2019SR0350697). Information regarding tumor size, density, and location on CT images was gathered from the Picture Archiving and Communication System (PACS) in our hospital (Carestream Health, Inc. 2023 (01) 60,889,971,048,875 (10) V12.1.5.6). Overall and LC-specific mortality data were obtained from the institutional information system of our institution (iMedical HIS R8.4.2, Electronic Medical Record System V4.5) and/or via regular telephone follow-up. Survival time was calculated from the date of pathological diagnosis confirmation to the date of death for nonsurvivors, the date of the first loss of follow-up for patients who missed three telephone follow-ups, or until Jan 31th, 2024, for survivors.

Statistical Analysis

All analyses were conducted using IBM SPSS 26.0 and the R package ggpubr (version: 4.2.1, survival [3.3.1], survminer [0.4.9], ggplot2 [3.3.6]). All the statistical tests were two-sided, and a p value < 0.05 was considered to indicate a statistically significant difference.

The baseline characteristics of the study population are described as frequencies (n) with proportions (%) for categorical variables and means with SDs for continuous variables. Continuous variables were compared with independent samples t tests or nonparametric tests, while categorical variables were compared with chi-square tests. For the primary outcomes of all-cause mortality and LC-specific mortality, the Kaplan‒Meier method was used to estimate cumulative incidence, and Cox regression and log-rank tests were used to assess differences between groups.

To estimate the association between risk factors and death in pure-solid stage T1 LC patients, two steps were taken. First, least absolute shrinkage selection operator (LASSO) analysis was performed for the population at risk, with death as the dependent variable and baseline characteristics, including age, smoking history, history of COPD, personal history of tumor, environmental exposure, family history, size of pulmonary nodules, pathological type (adenocarcinoma or not), micropapillary histologic pattern, and TNM stage (0-I or stage II-IV), as independent variables. Next, potential imbalanced factors were adjusted using a multivariate Cox regression model, and individual weights were applied to the models to estimate the weighted associations between risk factors and outcomes. Hazard ratios (HRs) and 95% confidence intervals (CIs) are reported.

Results

Baseline Characteristics

A total of 1876 patients with definite pathological diagnoses of stage T1 LC were enrolled. Of these, 1491 (79.48%) non-high-risk patients were included in the observation group, and the remaining 385 (20.52%) high-risk patients composed the control group (Fig. 1). For each year between 2019 and 2023, the proportion of patients with stage T1 LC categorized as non-high-risk (the observation group: 80.30%, 79.86%, 78.05%, 76.33%, and 81.06%, respectively) consistently exceeded that of the high-risk group (the control group: 19.70%, 20.14%, 21.95%, 23.67%, and 18.94%, respectively) (Fig. 2).

Fig. 1
figure 1

Flow diagram of patient enrollment

Full size image
Fig. 2
figure 2

The proportion of Non-high-risk and high-risk stage T1 LC patients enrolled each year from 2019 to 2023, by USPSTF criteria

Full size image

In the observation group, the mean age and tumor size were younger and smaller, respectively, than those in the control group (58.20 ± 12.29 years vs. 63.85 ± 7.18 years, 15.94 ± 6.86 mm vs. 19.22 ± 6.53 mm). Additionally, there were significantly higher proportions of females, subsolid LC, LUAD, and stage 0-I LC in the observation group than in the control group (72.84% vs. 0.52%, 62.11% vs. 41.30%, 63.58% vs. 27.79%, 93.90% vs. 71.17%, 84.17% vs. 62.60%, respectively). All p values were < 0.001 (Table 1).

Table 1 Comparison of baseline characteristics of stage T1 LC between non-high-risk and high-risk groups
Full size table

Primary outcomes

The primary analysis results indicated that the overall survival probability for patients in the observation group was significantly higher than that for patients in the control group (HR = 0.23, [95% CI: 0.18, 0.31], p < 0.001). Similarly, the LC-specific survival probability for patients in the observation group was also significantly higher than that for patients in the control group (HR = 0.23, [95% CI: 0.17, 0.31], p < 0.001). This conclusion was consistent across both the Cox regression and log-rank tests, two distinct statistical methods (Fig. 3a, a1, Fig. 4a, a1).

Fig. 3
figure 3

Risk factors independently associated with overall survival in pure-solid stage T1 LC patients by multivariate cox regression analysis

Full size image
Fig. 4
figure 4

Cumulative survival curves of stage T1 LC patients, by cox regression

Full size image

Subgroup analysis

In the pure-solid stage T1 LC subgroup, the proportions of LUAD and stage 0-I tumors in the observation group (84.90%, 56.91%) were higher than those in the control group (61.15%, 49.64%, p < 0.001, p < 0.05), although there was no significant difference in tumor size between the two groups (20.04 ± 6.17 mm, 20.63 ± 6.18 mm, p > 0.05) (Table 2). In the pure-solid stage Ia LC surgery subgroup, the proportion of LUAD patients in the observation group was greater than that in the control group (84.90% vs. 61.15%, p < 0.001). There were no noticeable differences in the mean size of the nodules, surgical approach, or occurrence of severe postoperative complications between the observation and control groups (all p > 0.05) (Table 3).

Table 2 Comparison of baseline supplementary characteristics of stage T1 LC in subgroups between non-high-risk and high-risk groups
Full size table
Table 3 Comparison of baseline characteristics of pure-solid stage Ia LC surgery patients between non-high-risk and high-risk groups
Full size table

The overall survival probability and the LC-specific survival probability of the pure-solid stage T1 LC subgroup (HR = 0.43, [95% CI: 0.33, 0.57], p < 0.001; HR = 0.44, [95% CI: 0.33, 0.60], p < 0.001), the pure-solid stage Ia LC surgery subgroup (HR = 0.29, [95% CI: 0.16, 0.52], p < 0.001; HR = 0.25, [95% CI: 0.09, 0.65], p = 0.005), and the pure-solid stage Ia LUAD surgery subgroup (HR = 0.24, [95% CI: 0.12, 0.48], p < 0.001; HR = 0.21, [95% CI: 0.07, 0.64], p = 0.006) were notably higher in the observation group than in the control group (Fig. 3c-d, 4c-d).

Multivariate Cox regression analysis indicated that older age (HR = 1.039, [95% CI: 1.023, 1.055], p < 0.001), smoking history (HR = 1.58, [95% CI: 1.176, 2.125], p < 0.05), adenocarcinoma status (HR = 1.424, [95% CI: 1.04, 1.949], p < 0.05), and advanced TNM stage (HR = 6.305, [95% CI: 4.613, 8.616], p < 0.001) were risk factors significantly associated with the death of patients with pure-solid stage T1 LC (area under the curve, AUC = 0.89) (Fig. 5, Fig. S1).

Fig. 5
figure 5

Cumulative survival curves of stage T1 LC patients, by Logrank Test

Full size image

Discussion

This is one of the few studies revealed that non-high-risk population with incidental stage T1 LC who did not meet the USPSTF criteria was younger, had smaller tumors, had fewer underlying diseases, had a greater proportion of subsolid nodules and LUAD, and had a lower probability of metastasis. The overall and LC-specific survival probabilities for the non-high-risk population were also higher than those for the high-risk population. These results appeared to be consistent across important subgroups.

LC screening of high-risk individuals cannot prevent most deaths [22, 23]. The detection rate of LC among non-high-risk individuals is increasing. In our study, the proportion of patients with incidental stage T1 LC categorized as non-high-risk (79.48%) who did not meet the USPSTF criteria consistently exceeded that of the high-risk group, which is consistent with the findings of a previous study. In 2022, Jing Ren et al. reported that for 1368 non-high-risk participants who did not meet the Chinese Expert Consensus on Diagnosis and Treatment of Pulmonary Nodules (2018 Edition) criteria, 264 LCs were identified, with an LC detection rate of 19.30%, which was significantly higher than that in the high-risk group (15.73%) [24]. The significantly high proportion of non-high-risk individuals with stage T1 LC observed in this study can be attributed to the high proportion of women included in our study from China, and all patients belonged to the population of incidental PNs.

Smoking history is a crucial factor in distinguishing between non-high-risk and high-risk individuals. Nonsmokers are those who have smoked fewer than < 100 cigarettes in their lifetime [25]. Several studies have shown that compared with smokers, LCINs are more likely to be female, to be Asian, to have a lower comorbidity burden and to have LUAD [26,27,28,29]. LUAD in never smokers significantly differs from that in smokers regarding both age at diagnosis and risk of lymph node metastasis [30]. However, there are limited studies on non-high-risk populations, including those who seldom smoke and those who smoked frequently in the past but who quit smoking many years ago. Jing Ren et al. reported that in the non-high-risk group, the percentages of patients with active smoking, passive smoking, a history of chronic comorbidities, a family history of LC, and occupational exposure were all significantly lower than those in the high-risk Group [24]. These findings were consistent with our results. However, in their study, there were no significant differences in pathological type between the non-high-risk and high-risk Groups [24]. In our study, we observed that non-high-risk patients with stage T1 LC were more likely to have LUAD than were high-risk patients. Radiologically subsolid nodules are mostly LUAD or adenocarcinoma precursors. In their study, the majority of malignant nodules in their cohorts were subsolid, comprising 78.03% in the non-high-risk group versus 79.37% in the high-risk group. However, in our study, there was a notable disparity in tumor density between the non-high-risk and high-risk groups, with pure-solid nodules accounting for 36.42% of the non-high-risk group and 72.21% of the high-risk group. The variance in pathological types may primarily be associated with the proportions of LCs with different densities.

The objective of diagnosing and treating LC is to decrease mortality. A low-density LC is usually regarded as an indolent LC considering its stable biological behavior. [31] Analyses of the mortality risk factors for pure-solid stage T1 LC in our study indicated that the impact of staging on LC prognosis is particularly significant. A series of studies have shown that screening with LDCT could help to detect LC at an earlier stage, and over 80% of these patients were confirmed to have LC at clinical stage I 3. Currently, in Denmark, 85% of cases of very early LC or stage IA LC originate from the IPN, which highlights the need for appropriate follow-up [32]. In this study, stage 0-I LC accounted for 79.74% of all patients with incidental stage T1 LC we enrolled, which was consistent with other reports. Studies in recent years have reported a type of highly invasive LC with unstable biological behaviors among pure-solid stage T1 LCs, which is characterized by local or distant metastasis even though the primary lesions still contain PNs. Through our subgroup analysis, we found that the proportion of stage 0-I LC patients in the subsolid stage T1 LC subgroup reached 99.43%, while that in the pure-solid stage T1 LC subgroup reached only 54.45%, which meant that nearly half of the pure-solid malignant nodules were not discovered until they advanced. We should emphasize this kind of highly invasive malignant nodule because it is one of the greatest challenges we are now facing in the screening, diagnosis and treatment of LC, and in the future, it may develop into the majority of advanced LC, thus becoming the principal cause of death in patients with stage T1 LC. [33, 34] In addition, 92.82% of the detected LCs were in stage I. Specifically, in the non-high-risk group, 92.05% were classified as stage I, while 94.44% in the high-risk group were in stage I according to Jing Ren's study. The clinical stages of the two groups were similar. [24] However, in this study, non-high-risk patients with incidental stage T1 LC were more likely to be at earlier stages than high-risk patients were. In particular, among patients with pure-solid stage T1 LC, the proportion of patients with stage 0-I LC in the non-high-risk population (56.91%) was still notably higher than that in the high-risk group (49.64%) when there was no marked difference in the size of primary tumors between the two populations, which implied that non-high-risk populations tended to have a lower probability of metastasis. Although the overall proportion of patients with advanced stage T1 LC in our study could not reflect the real situation completely, prospective multicenter investigations may be necessary to explore the real proportion of patients with this type of cancer among all patients with incidental stage T1 LC and to achieve the goal of early diagnosis of LC.

One study reported that among the 41,262 patients with NSCLC involved in a population-based cohort study, 11% were never smokers, and the results of the study showed that these patients had a higher general survival rate. [35] However, the multivariate analysis of a retrospective study, which enrolled over 4500 patients with NSCLC who had undergone surgeries (724 never smokers and 3822 current smokers), could not prove that smoking was an important determinant of overall survival (OS). [36] There is currently no evidence documenting the prognosis of non-high-risk patients with incidental stage T1 LC. We pioneered the study of outcomes of the non-high-risk population with incidental stage T1 LC and found that the 5-year overall survival probability and LC-specific survival probability in the non-high-risk population were both noticeably higher than those in the high-risk population. In addition, to reduce the influence of confounding factors on outcome, we compared patients with nodules of similar size in the pure-solid stage Ia LUAD surgery subgroup and concluded that the OS probability and LC-specific survival probability in the non-high-risk population were both markedly higher than those in the high-risk population, which suggested that the better prognosis was not only related to adenocarcinoma histology and earlier stages but also probably related to better biological behaviors of this kind of LC. [32] In addition to the confirmed relationship between the cumulative risk of smoking and age and the pathogenesis of LC, it is essential to explore whether the risk is tightly connected with biological behaviors in the future, which is critical to the precise prevention of LC.

Limitations

This study has several limitations. First, this was a single-center study in which the enrollees were all from the PNs Clinic of the Department of Respiratory Medicine, our institution, in which the high proportion of advanced pure-solid stage T1 LCs could not thoroughly reflect the real conditions of present-stage T1 LCs. However, the current situation exists in reality. Additionally, we established a non-high-risk group and a high-risk group to solve this problem, thus avoiding impacts on the main conclusion of this study. Second, the study lacked women in the high-risk population, which might have resulted in a shift in population composition. This phenomenon could be explained by the extremely low proportion of women smokers in China and the even lower percentage of women who could meet the smoking criteria for inclusion in the high-risk population. Third, the factors influencing the prognosis of patients with LC are complex, and information about mutations in cancer driver genes and the subsequent therapeutic course was especially difficult to obtain through follow-up in our observational study. [35] We cannot exclude the possibility of differences in prognosis caused by different treatments. However, we compared the survival probability of patients in the pure-solid stage Ia LUAD surgery subgroup and reached a similar conclusion.

Conclusions

Our study revealed that the non-high-risk population with incidental stage T1 LC who did not meet the USPSTF criteria was younger, had smaller tumors, had fewer underlying diseases, had a greater proportion of subsolid nodules and LUAD, and had a lower probability of metastasis. The overall and LC-specific survival probabilities for the non-high-risk population were also higher than those for the high-risk population. The results appeared to be consistent across important subgroups. Early diagnosis and treatment of incidental non-high-risk pure-solid pulmonary nodules who do not meet the USPSTF criteria are important for minimizing delays in diagnosis from LC patients.