Natural course of lung adenocarcinoma manifesting as ground-glass nodules: invasiveness assessment based on growth evaluation

Introduction

Lung cancer remains the primary contributor to global cancer-related morbidity and mortality (1). Among lung cancers, lung adenocarcinoma (LUAD) is the most prevalent pathological subtype, accounting for approximately 40% of all cases (2). With the increasing use of computed tomography (CT) screening for lung cancer and increased number of incidental CT findings, a form of early-stage LUAD, known as LUAD manifesting as ground-glass nodules (gLUAD), is being detected more frequently (3-5). Consequently, the clinical management of gLUAD has emerged as a critical area of focus.

Radiologically, gLUAD appears as a hazy area of increased density that does not obscure underlying structures such as lung vessels and bronchi (6). Based on the presence or absence of a solid component, gLUAD can be further classified into LUAD with pure ground-glass nodules (pgLUAD) or with mixed ground-glass nodules (mgLUAD). Pathologically, gLUAD is categorized into adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC) (7), with invasiveness increasing sequentially across these subtypes, respectively. In contrast to solid LUAD, gLUAD demonstrates distinct clinical characteristics (6,8-10). For one, gLUAD exhibits an indolent growth pattern, often necessitating prolonged follow-up (5 years or more) for definitive characterization. For another, although the overall prognosis of gLUAD is favorable, significant differences exist among its pathological subtypes. The 10-year disease-free survival (DFS) rate of patients with AIS/MIA is 100% following complete resection (11). However, for early-stage IAC, the 5-year DFS is only 89.0% (12). Furthermore, different pathological subtypes of gLUAD require distinct clinical management strategies (13,14). Patients with AIS/MIA are typically candidates for sublobar resection, with the timing of surgery determined through follow-up, whereas IAC is generally treated with immediate lobectomy, depending on the size of solid component. Therefore, understanding the growth patterns of gLUAD and assessing its invasiveness based on these characteristics are critical for optimizing follow-up and intervention strategies.

Previous studies on the growth process of gLUAD have primarily used a limited set of methods to evaluate its growth, without further investigation into the differences in the values of various evaluation methods for assessing invasiveness (15-18). This study aimed to systematically investigate the growth dynamics of gLUAD and compare the values of different growth evaluation methods in predicting its invasiveness. We present this article in accordance with the STARD reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-395/rc).

Methods

Study participants

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective study was approved by the ethics committee of Chinese PLA General Hospital (No. S2023-206-01) and was registered at ChiCTR.org.cn (ChiCTR2300072039). Informed consent was waived due to the retrospective nature of the study.

Participants were retrospectively enrolled from the First and Fourth Medical Centers of the PLA General Hospital between January 2018 and December 2022 (Figure 1). The inclusion criteria were as follows: (I) pathologically confirmed LUAD and (II) a follow-up duration exceeding 3 months prior to intervention. Meanwhile, the exclusion criterion was fewer than three available lung CT scans (slice thickness ≤1.5 mm).

Figure 1 Participant selection flowchart. Center 1, the First Medical Center of the Chinese PLA General Hospital; Center 2, the Fourth Medical Center of the Chinese PLA General Hospital. AIS, adenocarcinoma in situ; CT, computed tomography; gLUAD, lung adenocarcinoma manifesting as ground-glass nodules; IAC, invasive adenocarcinoma; MIA, minimally invasive adenocarcinoma.

A total of 625 gLUADs from 564 participants were included in this study. The cohort consisted of 341 female (60.5%) and 223 male (39.5%) participants, with a median age of 57.0 years.

Image acquisition

CT scans were acquired using Siemens Healthineers (Erlangen, Germany) or Philips (Andover, the Netherlands) scanners, with tube voltages typically set at 100 or 120 kVp. The images were reconstructed with a matrix size of 512×512, a slice thickness of 1.0–1.5 mm, and a pixel spacing of 0.7–1.0 mm. All images used in this study were unenhanced.

Lesion measurement

A senior pulmonologist (Z.Y.) manually segmented the total lesion and its solid components (if present) using 3D Slicer version 5.6.1 (slicer.org). The pulmonologist was blinded to the pathological results. Lesion size, volume, and mean CT values were computed with PyRadiomics version 3.1.0 (radiomics.io) (19). The lesion mass was computed using the following formula:

Mass=meanCT+10001000⋅Volume[1]

To evaluate the inter-observer agreement of manual segmentation-based measurements, three physicians (Z.Y., Y.W., and J.R.) independently conducted duplicate lesion measurements on an additional small dataset (n=100). The results demonstrated satisfactory inter-observer agreement, with intraclass correlation coefficients (ICCs) of 0.98/0.97 for total/solid size, 0.99/0.97 for volume, and 0.99/0.97 for mass (see Table S1 for additional data). The measurement variations in our manual measurements were comparable to those reported in relevant studies, meeting current benchmark standards for this methodology (20,21).

Growth pattern analysis

To analyze lesion growth patterns, we modeled the lesion growth using linear and exponential growth assumptions.

Under the linear growth assumption, lesion growth was described by the following equation:

d=αL⋅t+βL[2]

where d is the lesion size, αL is the linear growth rate (LGR), and βL is the initial lesion dimension.

Under the exponential growth assumption, lesion growth was described by the following equation:

d=2αE⋅t+βE[3]

where d is the lesion dimension, αE is the exponential growth rate (EGR), and βE is a constant related to the initial lesion dimension. The EGR and doubling time are inversely related.

Adjusted R² and the root mean squared error (RMSE) were computed to evaluate the model fit: the higher the R² and the lower the RMSE are, the better the model fit.

Growth evaluation criteria

This study applied four definitions to determine whether a lesion exhibited growth: size growth, volume growth, and stage shift. Size growth was defined as an increase of at least 2 mm in the total size or the size of the solid component, or the appearance of a new solid component (22). Volume growth was defined as an increase in the total volume by at least 25% (23). Mass growth was defined as an increase in the total volume by at least 30% (24). Stage shift was defined as upstaging of the lesion’s clinical classification based on the ninth edition of the tumor-node-metastasis (TNM) classification and staging system (25) (Figure S1).

Statistical analysis

Statistical analyses were conducted using R version 4.4.1 (The R Foundation for Statistical Computing, Vienna, Austria; r-project.org) and Python version 3.11.5 (Python Software Foundation, Wilmington, DE, USA; python.org). Categorical variables were compared between groups using the Chi-squared test. For continuous variables, normality was assessed with the Kolmogorov-Smirnov test. The Student’s t-test was applied when normality assumptions were met; otherwise, the Mann-Whitney test was used. The Kaplan-Meier (KM) curves for different groups were compared using the log-rank test. Differences in the area under the curve (AUC) between indicators were evaluated using the DeLong test. A P value <0.05 was considered statistically significant. Confidence intervals (CIs) for assessment metrics were calculated using the bootstrap method with 1,000 resamples.

Results

gLUAD characteristics

The lesions included 383 (61.3%) mgLUAD and 242 (38.7%) pgLUAD, with a median initial size of 5.53 mm and a median follow-up duration of 1,050 days. The pathological diagnoses included AIS (43, 6.88%), MIA (139, 22.2%), and IAC (443, 70.9%). Significant differences were observed between AIS/MIA and IAC for several characteristics, including mgLUAD, lobulation, spiculation, vascular convergence, consolidation-to-tumor ratio (CTR), total/solid size, total/solid volume, total/solid mass, clinical T staging, size growth, volume growth, mass growth, and stage shift. These lesion characteristics are summarized in Table 1.

Growth pattern analysis

gLUAD growth was modeled using linear and exponential growth hypotheses (Figure 2). The results indicated that both models demonstrated similar fitting performance for gLUAD growth trajectories. The linear growth model provided satisfactory fit for total/solid size, volume, and mass, with adjusted R² values of 0.98/0.95, 0.98/0.90, and 0.95/0.90, respectively, and RMSE values of 0.53/0.92, 169.80/100.00, and 116.12/101.24, respectively; similarly, the exponential growth model achieved adjusted R² values of 0.98/0.95, 0.98/0.93, and 0.98/0.92, with corresponding RMSE values of 0.51/0.94, 143.87/106.81, and 72.81/108.99, respectively. These findings justify the application of both linear and exponential growth models to describe gLUAD growth.

Figure 2 Fitting of the linear and exponential models to growth. (A) Total size. (B) Solid size. (C) Total volume. (D) Solid volume. (E) Total mass. (F) Solid mass. The dots of different colors represent the measurement data and estimated outcomes of distinct cases. The values in the parentheses represents 95% confidence interval. RMSE, root mean squared error.

Growth evaluation

The growth trajectories of individual gLUAD, along with the median growth trajectories of all cases of gLUAD, AIS/MIA, and IAC estimated using the exponential growth model, are shown in Figure 3A-3F. Overall, gLUAD exhibited a growth trend over time, but differences were observed in the total/solid size, volume, and mass growth trajectories between AIS/MIA and IAC. Compared to the AIS/MIA, IAC demonstrated significantly greater growth. Two gLUAD cases were selected as examples to illustrate these differences in growth trajectories: the first was an AIS from a 50-year-old female, and the second was an IAC from a 54-year-old female (Figure 3G-3J).

Figure 3 Growth trajectories and follow-up case examples. (A) Total size. (B) Total volume. (C) Total mass. (D) Solid size. (E) Solid volume. (F) Solid mass. The solid lines in different colors represent the growth trajectories of distinct cases. (G,H) Follow-up imaging and growth trajectories of an AIS case. (I,J) Follow-up imaging and growth trajectories of an IAC case. AIS, adenocarcinoma in situ; IAC, invasive adenocarcinoma; MIA, minimally invasive adenocarcinoma.

gLUAD growth was qualitatively evaluated based on the four defined growth criteria (Table 2). KM plots for different types of gLUAD growth are presented in Figure 4. Significant differences were observed between AIS/MIA and IAC in size growth, volume growth, mass growth, and stage shift (all P values <0.001). The cumulative probabilities of size growth, volume growth, mass growth, and stage shift within 1 year for gLUAD were 0.13, 0.23, 0.18, and 0.11, respectively; for the AIS/MIA type, these probabilities were 0.10, 0.16, 0.09, and 0.10, respectively, while for the IAC subtype, they were 0.14, 0.26, 0.21, and 0.12, respectively. The median growth times for size growth, volume growth, mass growth, and stage shift in gLUAD were 1,273, 750, 792, and 1,672 days, respectively; for AIS/MIA, the median growth times were 1,902, 1,175, 1,569, and 2,229 days, respectively, while for IAC, they were 1,113, 681, 730, and 1,420 days, respectively. Additionally, after 5 years of follow-up, gLUAD still had an approximately 0.1 probability of remaining stable (i.e., showing no growth based on any of the defined criteria). This probability was around 0.2 for AIS/MIA and about 0.05 for IAC.

Figure 4 Kaplan-Meier plots of growth. (A) Size growth. (B) Volume growth. (C) Mass growth. (D) Stage shift. AIS, adenocarcinoma in situ; IAC, invasive adenocarcinoma; MIA, minimally invasive adenocarcinoma.

The gLUAD growth rates were used to quantitatively evaluate lesion growth (Table 3 and Figure S2). Based on the linear growth model, the median growth rates of total/solid size, volume, and mass for gLUAD were 0.41/0.25 mm per year, 80.5/1.87 mm³ per year, and 31.9/1.72 mg per year, respectively. Based on the exponential growth model, the median growth rates of total/solid size, volume, and mass were 0.09/0.12 per year, 0.25/0.36 per year, and 0.24/0.36 per year, corresponding to median doubling times of >3,650/3,042, 1,460/1,014, and 1,521/1,014 days, respectively. For AIS/MIA, all doubling times exceeded 3,650 days, whereas for IAC, they were 3,318/2,147 days, 1,141/777 days, and 1,074/760 days, respectively. Regardless of the growth model used, significant differences in growth rates of total/solid size, volume, and mass were observed between AIS/MIA and IAC (all P values <0.001).

Furthermore, risk factors associated with gLUAD growth were analyzed (Tables S2-S5). Total size was an independent risk factor for size growth, spiculation was an independent risk factor for volume and mass growth, and sex was an independent risk factor for all four definitions of growth.

Invasiveness assessment based on growth evaluation

The receiver operating characteristic (ROC) curves for various growth evaluation methods in differentiating pathological subtypes of gLUAD are presented in Figure 5A-5C, and the corresponding metrics are summarized in Table 4.

Figure 5 ROC curves for invasiveness assessment and predictive evaluation of the LGR of total mass. ROC curves for growth definitions (A), LGRs (B), and EGRs (C). Predictive evaluation of total mass LGR: calibration plot (D), decision curve analysis (E), and confusion matrix (F). AIS, adenocarcinoma in situ; AUC, area under the curve; EGR, exponential growth rate; IAC, invasive adenocarcinoma; LGR, linear growth rate; MIA, minimally invasive adenocarcinoma; ROC, receiver operating characteristic.

The AUC for differentiating gLUAD invasiveness using growth definitions ranged from 0.61 to 0.69. Specifically, assessments based on size growth and stage shift provided relatively higher specificity in identifying invasive gLUAD (sensitivity: 0.63/0.48; specificity: 0.67/0.75), whereas assessments based on volume growth and mass growth exhibited relatively higher sensitivity (sensitivity: 0.81/0.78; specificity: 0.60/0.75).

In assessing gLUAD invasiveness using qualitative growth analysis, LGRs demonstrated better discriminative ability for gLUAD, with a different degree of invasiveness as compared to EGRs (all P values <0.05). The AUCs for LGRs vs. EGRs of total/solid size, volume, and mass were 0.74/0.70 vs. 0.71/0.66 (P<0.001/<0.001), 0.79/0.73 vs. 0.76/0.68 (P=0.01/<0.001), and 0.83/0.72 vs. 0.79/0.68 (P=0.002/0.001), respectively.

Among all methods, the LGR of total mass exhibited the highest discriminative ability, with an AUC of 0.83. At the optimal cutoff value (16.80 mg/year), the LGR of total mass achieved an accuracy of 0.78, a sensitivity of 0.77, and a specificity of 0.79 in distinguishing the invasive subtypes of gLUAD. We additionally calculated the risk probability corresponding to the LGR of total mass using logistic regression. The calibration curve, clinical decision curve, and confusion matrix are shown in Figure 5D-5F, respectively.

Discussion

Lung cancer is a major contributor to the cancer-related health burden, with LUAD being its predominant pathological subtype. With the widespread use of CT screening and an increasing number of CT examinations, the frequency of gLUAD being detected is rising. gLUAD generally exhibits indolent growth and has a favorable prognosis. Different pathological subtypes require distinct clinical management strategies. Our study systematically analyzed the growth patterns of gLUAD and examined invasiveness assessment based on growth characteristics, which may provide valuable insights for optimizing follow-up and intervention strategies in clinical practice.

The intrinsic properties of tumors and tumor-host interactions determine the growth process of lung cancer. For example, a study found that the growth of gLUAD is associated with its pathological subtype and Ki-67 expression (26). Research into the growth characteristics of lung cancer has provided valuable insights into its underlying dynamic biological processes and can guide clinical management. Recent studies on lung cancer growth characteristics indicate that lung cancer growth is a complex process, and conclusions regarding its growing patterns vary (27-30). Some studies suggest an exponential growing pattern, while others propose an accelerated growth pattern. Additionally, some research indicates that lung cancer growth does not conform to common patterns such as linear, exponential, or Gompertzian, a well-established biological growing pattern (31). In clinical practice, the exponential growth model is more commonly used to describe lung cancer progression. For gLUAD, a previous study suggested that among linear, quadratic, and power-law models, the exponential model provided a slightly better fit (30). However, that study also highlighted the heterogeneity of gLUAD growth and pointed out that the exponential model was not a perfect representation. In our study, a substantial proportion of gLUAD had only 3 longitudinal growth measurements available. The application of complex models to these data carries inherent risks of overfitting and numerical instability. For instance, the Gompertzian model theoretically requires at least 4 time-point measurements to achieve robust fitting. Therefore, we strategically limited our analysis to more robust linear and exponential growth models for characterizing gLUAD growth dynamics. Our findings showed comparable fitting performance between the two models. One potential explanation for this is that due to the indolent growth of gLUAD, the exponential model can be approximated as a linear model over a limited period. Mathematically, when αE⋅t is small, the exponential model d=2αE⋅t+βE can be approximated as a linear model d≈(2βE⋅ln2⋅αE)⋅t+2βE. Our results support the ability of the linear model to describe gLUAD growth adequately. Specifically, for total volume, the exponential model demonstrated a slight advantage over the linear model, which aligns with previous research (30).

Fleischner guidelines recommend terminating follow-up if a ground-glass nodule remains stable for 5 years (32). However, a recent expert consensus recommends that even stable ground-glass nodules should undergo extended-interval follow-up (6). Our study found that within 5 years of follow-up, gLUAD had an approximately 10% probability of showing no growth under any definition, even among cases of invasive gLUAD, where this probability remained at around 5%. Our findings support the recent consensus advocating for extended follow-up of stable ground-glass nodules given their high growth potential. A recent study from Korea also reported that approximately 5% of stable ground-glass nodules after 5 years of follow-up exhibit growth during subsequent follow-ups, with more than half of these nodules even undergoing stage shift (18). This finding is consistent with the conclusions of our study. In addition, surgical resection represents the principal therapeutic intervention for gLUAD. However, the optimal timing of resection poses a critical clinical issue (33). Our findings indicate that the therapeutic window for gLUAD is relatively wide, and the most appropriate timing should be individualized based on a comprehensive assessment of the patient’s circumstances. For example, early surgical intervention may be warranted in high-risk patients experiencing diagnosis-related refractory anxiety, as prompt resection could enhance quality of life by alleviating disease-related distress (34). Conversely, in cases where low-risk patients are navigating pivotal personal or professional commitments, deliberate deferral of surgery is unlikely to result in significant disease progression and may better align with their long-term priorities.

Many previous studies have examined risk factors for ground-glass nodule growth. Multiple meta-analyses have identified gender, age, smoking history, personal lung cancer history, lesion size, volume, lobulation, spiculation, and vacuolation as potential risk factors (35-37). In our multivariate analysis, gender, lesion size, and spiculation were significant independent predictors of gLUAD growth, whereas other factors showed no significant predictive value. This discrepancy may be partially attributed to differences in study populations. Our study specifically included pathologically confirmed gLUAD, whereas previous studies analyzed all ground-glass nodules, potentially including benign lesions and a higher proportion of noninvasive gLUAD. Regarding gLUAD growth rates, our findings confirmed that gLUAD exhibited indolent progression. The median doubling times for total/solid size, volume, and mass were >3,650/3,042, 1,460/1,014, and 1,521/1,014 days, respectively. Compared to AIS/MIA, IAC exhibited significantly faster growth rates, with median doubling times of 3,318/2,147, 1,141/777, and 1,074/760 days for total/solid size, volume, and mass, respectively. These findings are consistent with previous studies (16,17,21).

Predicting the invasiveness of gLUAD based on follow-up data is a crucial clinical challenge. Several studies have developed dynamic radiomics-based models for gLUAD invasiveness prediction (38-42). These studies have leveraged advanced machine learning algorithms and achieved promising results. However, due to the “black-box” nature of these algorithms and their high computational demands, clinical applicability remains limited in terms of interpretability and feasibility. In contrast, growth evaluation-based invasiveness assessment offers better interpretability and feasibility for clinical use. Our study compared various qualitative and quantitative growth evaluation methods for predicting gLUAD invasiveness. The results indicated that qualitative growth evaluation had limited predictive value (AUC: 0.61–0.69). Among these methods, volume growth and mass growth assessments demonstrated higher sensitivity than did size growth and stage shift; however, this came at the expense of specificity. This may be due to the superior ability of volume growth and mass growth to capture lesion growth (43,44). For quantitative growth evaluation, an unexpected finding was that LGRs outperformed EGRs in predicting gLUAD invasiveness. We hypothesize that this phenomenon may be attributed to greater heterogeneity in LGRs among gLUAD with different invasiveness. In accordance with our theoretical framework, for most cases, the differences in LGRs and EGRs between AIS/MIA and IAC exhibit the following approximate relationship: ΔαL≈2βE⋅ln2⋅(ΔαE+ln2⋅αE⋅ΔβE)>ΔαE. The LGR of total mass exhibited the highest discriminatory ability, with an AUC of 0.83. At the optimal cutoff value, it demonstrated improved diagnostic performance compared with the EGR of total mass, achieving an accuracy of 0.78 (vs. 0.70), sensitivity of 0.77 (vs. 0.65), while maintaining comparable specificity (0.79 vs. 0.81). We hypothesize that total mass is the best discriminator of gLUAD invasiveness because it reflects more comprehensively the gLUAD growth. gLUAD progression involves both an increase in dimension and alveolar space filling, and total mass likely captures these two processes more effectively. For the clinical management of gLUAD, our findings suggest that the LGR of total mass may serve as a superior invasion assessment metric compared to the conventional EGR of total mass (corresponding to mass doubling time). In clinical applications, the LGR of total mass may serve as a predictive biomarker for invasiveness risk in gLUAD, facilitating personalized follow-up strategies and surgical timing optimization. Our findings demonstrate that gLUAD with an LGR of total mass ≥16.80 mg/year is highly suggestive of IAC with fast-growing potential. These cases warrant more vigilant CT surveillance intervals and should be considered for earlier surgical intervention. Conversely, gLUAD demonstrating an LGR of total mass <16.80 mg/year predominantly correlates with AIS or MIA, for which conservative CT follow-up represents the preferred management approach to mitigate risks of overtreatment.

Our study involved certain limitations which should be noted. First, only pathologically confirmed gLUAD meeting specific follow-up time and frequency criteria were included. This introduced potential selection bias. Second, as a retrospective study, variability in CT scan protocols and preoperative follow-up strategies could have introduced uncertainties. Third, lesion measurements involved manual input, which, despite being the current gold standard, inevitably introduced a degree of subjectivity. Fourth, our investigation into the growth patterns of gLUAD primarily relied on mathematical modeling. However, the conclusions and interpretations may require further validation through studies elucidating the underlying biological mechanisms. Finally, our study data were derived exclusively from two tertiary medical centers in Beijing, China. While these institutions serve clinically diverse patient populations, the cohort predominantly comprised individuals of Han Chinese ethnicity from the Beijing region. Furthermore, institutional practices for managing gLUAD may not fully reflect the heterogeneity of clinical decision-making across other healthcare systems, particularly those with differing resource availability or practice guidelines. To mitigate potential biases, we employed statistical reinforcement techniques, including bootstrapping analyses, to enhance the robustness of our primary conclusions. Nevertheless, the generalizability of these findings remains constrained by the geographic and ethnic homogeneity of the study population. Therefore, the generalizability of our findings requires further validation. Given these limitations, future large-scale, multicenter prospective studies will be conducted to provide further validation of our findings.

Conclusions

gLUAD generally exhibits indolent growth while IAC shows a more pronounced growth trend as compared to AIS/MIA. Both linear and exponential growth models demonstrate a similar fitting performance for gLUAD growth, but the linear growth model offers a more reliable assessment of invasiveness. The LGR of total mass has greater predictive value for gLUAD invasiveness and may serve as a useful indicator in clinical assessment.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-395/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-395/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-395/prf

Funding: This work was supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2024ZD0529503/2024ZD052950) and the Military Scientific Research Project (No. 24BJZ08).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-395/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 and was approved by the ethics committee of Chinese PLA General Hospital (No. S2023-206-01). The requirement for individual consent was waived due to the retrospective nature of the analysis.

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. Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin 2024;74:12-49. [Crossref] [PubMed]
2. Leiter A, Veluswamy RR, Wisnivesky JP. The global burden of lung cancer: current status and future trends. Nat Rev Clin Oncol 2023;20:624-39. [Crossref] [PubMed]
3. Sadate A, Occean BV, Beregi JP, et al. Systematic review and meta-analysis on the impact of lung cancer screening by low-dose computed tomography. Eur J Cancer 2020;134:107-14. [Crossref] [PubMed]
4. Kim YW, Kwon BS, Lim SY, et al. Lung cancer probability and clinical outcomes of baseline and new subsolid nodules detected on low-dose CT screening. Thorax 2021;76:980-8. [Crossref] [PubMed]
5. Mazzone PJ, Lam L. Evaluating the Patient With a Pulmonary Nodule: A Review. JAMA 2022;327:264-73. [Crossref] [PubMed]
6. Chen H, Kim AW, Hsin M, et al. The 2023 American Association for Thoracic Surgery (AATS) Expert Consensus Document: Management of subsolid lung nodules. J Thorac Cardiovasc Surg 2024;168:631-647.e11. [Crossref] [PubMed]
7. Travis WD, Brambilla E, Nicholson AG, et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of Genetic, Clinical and Radiologic Advances Since the 2004 Classification. J Thorac Oncol 2015;10:1243-60. [Crossref] [PubMed]
8. Ye T, Deng L, Wang S, et al. Lung Adenocarcinomas Manifesting as Radiological Part-Solid Nodules Define a Special Clinical Subtype. J Thorac Oncol 2019;14:617-27. [Crossref] [PubMed]
9. Hammer MM, Hatabu H. Subsolid pulmonary nodules: Controversy and perspective. Eur J Radiol Open 2020;7:100267. [Crossref] [PubMed]
10. Lai J, Li Q, Fu F, et al. Subsolid Lung Adenocarcinomas: Radiological, Clinical and Pathological Features and Outcomes. Semin Thorac Cardiovasc Surg 2022;34:702-10. [Crossref] [PubMed]
11. Yotsukura M, Asamura H, Motoi N, et al. Long-Term Prognosis of Patients With Resected Adenocarcinoma In Situ and Minimally Invasive Adenocarcinoma of the Lung. J Thorac Oncol 2021;16:1312-20. [Crossref] [PubMed]
12. Watanabe Y, Hattori A, Nojiri S, et al. Clinical impact of a small component of ground-glass opacity in solid-dominant clinical stage IA non-small cell lung cancer. J Thorac Cardiovasc Surg 2022;163:791-801.e4. [Crossref] [PubMed]
13. Saji H, Okada M, Tsuboi M, et al. Segmentectomy versus lobectomy in small-sized peripheral non-small-cell lung cancer (JCOG0802/WJOG4607L): a multicentre, open-label, phase 3, randomised, controlled, non-inferiority trial. Lancet 2022;399:1607-17. [Crossref] [PubMed]
14. Suzuki K, Watanabe SI, Wakabayashi M, et al. A single-arm study of sublobar resection for ground-glass opacity dominant peripheral lung cancer. J Thorac Cardiovasc Surg 2022;163:289-301.e2. [Crossref] [PubMed]
15. Tang EK, Chen CS, Wu CC, et al. Natural History of Persistent Pulmonary Subsolid Nodules: Long-Term Observation of Different Interval Growth. Heart Lung Circ 2019;28:1747-54. [Crossref] [PubMed]
16. Kakinuma R, Noguchi M, Ashizawa K, et al. Natural History of Pulmonary Subsolid Nodules: A Prospective Multicenter Study. J Thorac Oncol 2016;11:1012-28. [Crossref] [PubMed]
17. Qi LL, Wang JW, Yang L, et al. Natural history of pathologically confirmed pulmonary subsolid nodules with deep learning-assisted nodule segmentation. Eur Radiol 2021;31:3884-97. [Crossref] [PubMed]
18. Lee JH, Lim WH, Park CM. Growth and Clinical Impact of Subsolid Lung Nodules ≥6 mm During Long-Term Follow-Up After Five Years of Stability. Korean J Radiol 2024;25:1093-9. [Crossref] [PubMed]
19. van Griethuysen JJM, Fedorov A, Parmar C, et al. Computational Radiomics System to Decode the Radiographic Phenotype. Cancer Res 2017;77:e104-7. [Crossref] [PubMed]
20. Kim H, Park CM, Hwang EJ, et al. Pulmonary subsolid nodules: value of semi-automatic measurement in diagnostic accuracy, diagnostic reproducibility and nodule classification agreement. Eur Radiol 2018;28:2124-33. [Crossref] [PubMed]
21. Song YS, Park CM, Park SJ, et al. Volume and mass doubling times of persistent pulmonary subsolid nodules detected in patients without known malignancy. Radiology 2014;273:276-84. [Crossref] [PubMed]
22. Bankier AA, MacMahon H, Goo JM, et al. Recommendations for Measuring Pulmonary Nodules at CT: A Statement from the Fleischner Society. Radiology 2017;285:584-600. [Crossref] [PubMed]
23. Callister ME, Baldwin DR, Akram AR, et al. British Thoracic Society guidelines for the investigation and management of pulmonary nodules. Thorax 2015;70:ii1-ii54. [Crossref] [PubMed]
24. Scholten ET, de Jong PA, de Hoop B, et al. Towards a close computed tomography monitoring approach for screen detected subsolid pulmonary nodules? Eur Respir J 2015;45:765-73. [Crossref] [PubMed]
25. Rami-Porta R. The TNM classification of lung cancer-a historic perspective. J Thorac Dis 2024;16:8053-67. [Crossref] [PubMed]
26. Huang S, Zhou H, Lin C, et al. The Correlation Between the Natural Course, Pathologic Properties With Ki-67 Expression in Lung Adenocarcinoma Presenting as Ground-Glass Nodules. Cancer Med 2024;13:e70390. [Crossref] [PubMed]
27. Ko JP, Berman EJ, Kaur M, et al. Pulmonary Nodules: growth rate assessment in patients by using serial CT and three-dimensional volumetry. Radiology 2012;262:662-71. [Crossref] [PubMed]
28. Heuvelmans MA, Vliegenthart R, de Koning HJ, et al. Quantification of growth patterns of screen-detected lung cancers: The NELSON study. Lung Cancer 2017;108:48-54. [Crossref] [PubMed]
29. Pérez-García VM, Calvo GF, Bosque JJ, et al. Universal scaling laws rule explosive growth in human cancers. Nat Phys 2020;16:1232-7. [Crossref] [PubMed]
30. de Margerie-Mellon C, Ngo LH, Gill RR, et al. The Growth Rate of Subsolid Lung Adenocarcinoma Nodules at Chest CT. Radiology 2020;297:189-98. [Crossref] [PubMed]
31. Tjørve KMC, Tjørve E. The use of Gompertz models in growth analyses, and new Gompertz-model approach: An addition to the Unified-Richards family. PLoS One 2017;12:e0178691. [Crossref] [PubMed]
32. MacMahon H, Naidich DP, Goo JM, et al. Guidelines for Management of Incidental Pulmonary Nodules Detected on CT Images: From the Fleischner Society 2017. Radiology 2017;284:228-43. [Crossref] [PubMed]
33. Zhang Y, Chen Z, Hu H, et al. Surgical Strategies for Pre- and Minimally Invasive Lung Adenocarcinoma 3.0: Lessons Learned From the Optimal Timing of Surgical Intervention. Semin Thorac Cardiovasc Surg 2022;34:311-4. [Crossref] [PubMed]
34. Li L, Zhao Y, Li H. Assessment of anxiety and depression in patients with incidental pulmonary nodules and analysis of its related impact factors. Thorac Cancer 2020;11:1433-42. [Crossref] [PubMed]
35. Liang X, Liu M, Li M, et al. Clinical and CT Features of Subsolid Pulmonary Nodules With Interval Growth: A Systematic Review and Meta-Analysis. Front Oncol 2022;12:929174. [Crossref] [PubMed]
36. Wu L, Gao C, Kong N, et al. The long-term course of subsolid nodules and predictors of interval growth on chest CT: a systematic review and meta-analysis. Eur Radiol 2023;33:2075-88. [Crossref] [PubMed]
37. Gao C, Li J, Wu L, et al. The Natural Growth of Subsolid Nodules Predicted by Quantitative Initial CT Features: A Systematic Review. Front Oncol 2020;10:318. [Crossref] [PubMed]
38. Tao G, Shi D, Yu L, et al. Longitudinal prediction of lung nodule invasiveness by sequential modelling with common clinical computed tomography (CT) measurements: a prediction accuracy study. Transl Lung Cancer Res 2022;11:845-57. [Crossref] [PubMed]
39. Singh A, Roshkovan L, Horng H, et al. Radiomics Analysis for the Identification of Invasive Pulmonary Subsolid Nodules From Longitudinal Presurgical CT Scans. J Thorac Imaging 2025;40:e0800. [PubMed]
40. Xu Y, Li Y, Yin H, et al. Consecutive Serial Non-Contrast CT Scan-Based Deep Learning Model Facilitates the Prediction of Tumor Invasiveness of Ground-Glass Nodules. Front Oncol 2021;11:725599. [Crossref] [PubMed]
41. Ma Y, Ma W, Xu X, et al. How Does the Delta-Radiomics Better Differentiate Pre-Invasive GGNs From Invasive GGNs? Front Oncol 2020;10:1017. [Crossref] [PubMed]
42. Li H, Luo Q, Zheng Y, et al. A nomogram for predicting invasiveness of lung adenocarcinoma manifesting as ground-glass nodules based on follow-up CT imaging. Transl Lung Cancer Res 2024;13:2617-35. [Crossref] [PubMed]
43. Zhang Z, Zhou L, Yang F, et al. The natural growth history of persistent pulmonary subsolid nodules: Radiology, genetics, and clinical management. Front Oncol 2022;12:1011712. [Crossref] [PubMed]
44. de Margerie-Mellon C, Gill RR, Monteiro Filho AC, et al. Growth Assessment of Pulmonary Adenocarcinomas Manifesting as Subsolid Nodules on CT: Comparison of Diameter-Based and Volume Measurements. Acad Radiol 2020;27:1385-93. [Crossref] [PubMed]