Segmentectomy versus lobectomy for stage I non-small cell lung cancer: a real-world propensity score-matched analysis

Introduction

While lobectomy remains the standard for resectable non-small cell lung cancer (NSCLC) since the Lung Cancer Study Group trial (1), increased detection of small tumors through computed tomography (CT) screening has renewed interest in parenchyma-sparing approaches (2,3).

Recent trials challenge this paradigm. JCOG0802/WJOG4607L demonstrated superior overall survival (OS) with segmentectomy for tumors ≤2 cm (4). while CALGB 140503 showed non-inferior disease-free survival (DFS) with sublobar resection (5). However, critical distinctions exist between anatomical segmentectomy—involving complete bronchopulmonary segment removal with systematic lymphadenectomy—and wedge resection (6,7). Since pivotal trials included substantial wedge resection proportions, anatomical segmentectomy’s true efficacy remains unclear (8,9).

Anatomical segmentectomy offers theoretical advantages: preserved parenchyma enables future interventions for second primaries (10,11) and maintains pulmonary reserve for systemic therapy tolerance (12,13). However, concerns persist regarding incomplete lymphatic clearance and technical challenges of intersegmental dissection (14-17).

This propensity-matched study compares anatomical segmentectomy versus lobectomy for stage I NSCLC, focusing exclusively on anatomical resection to clarify its role in modern early-stage lung cancer management. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1715/rc).

Methods

Study design and setting

This Retrospective cohort study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by the Wolfson Medical Center Institutional Review Board (No. 079-25-WOMC), using a prospectively maintained database at Wolfson Medical Center, Holon, Israel. The database was searched from January 2017 to December 2024 for stage I NSCLC pulmonary resections. The need for individual patient consent was waived due to the retrospective nature of the study.

Patient selection

From 178 NSCLC patients, 60 met the inclusion criteria: pathologically confirmed stage I NSCLC ≤3.5 cm [T1a–T2aN0M0, 8th American Joint Committee on Cancer (AJCC)]. Segmentectomy group: clinical T1a (≤2 cm) with intraoperative hilar sampling. Exclusion criteria included: neoadjuvant therapy, multiple primaries, wedge resections, incomplete staging, previous thoracic surgery (Figure 1).

Figure 1 Flow diagram of patient selection and allocation. The diagram shows the screening of 178 patients with NSCLC, exclusion criteria, and final propensity-matched cohort of 50 patients (25 segmentectomy, 25 lobectomy) included in the analysis. NSCLC, non-small cell lung cancer.

Surgical decision-making and technique

For tumors ≤2 cm: frozen section of ≥1 hilar node; if negative, anatomical segmentectomy performed per recent trial protocols (4). Anatomical segmentectomy defined as complete bronchopulmonary segment removal with segmental bronchus/artery/vein division and intersegmental plane dissection (5). For tumors >2 cm or based on surgeon preference (tumor location/patient factors): lobectomy without mandatory node sampling. Approach [video-assisted thoracoscopic surgery (VATS)/thoracotomy] determined by surgeon based on tumor/patient characteristics.

Lymph node assessment

All patients underwent radical systematic lymphadenectomy: ≥3 mediastinal stations plus hilar and intrapulmonary nodes. Right-sided: stations 2R, 4R, 7, 8, 9, 10R, 11R. Left-sided: stations 5, 6, 7, 8, 9, 10L, 11L, per International Association for the Study of Lung Cancer (IASLC) guidelines (18,19).

Data collection

Prospectively collected data included demographics, smoking history, comorbidities, pulmonary function [forced expiratory volume in 1 second (FEV₁)%, diffusing capacity for carbon monoxide (DLCO)%], American Society of Anesthesiologists (ASA) score, and tumor characteristics (size, location, histology, staging). Perioperative outcomes: operative time, blood loss, conversions, hospital stay, chest tube duration, and complications (Clavien-Dindo classification). No missing data were observed for any variable; all patients completed follow-up or died during the study period.

Primary and secondary endpoints

The primary endpoints were OS and DFS. OS was calculated from the date of surgery to death from any cause or last follow-up. DFS was defined as the time from surgery to first recurrence (local, regional, or distant) or cancer-related death. Secondary endpoints included patterns of recurrence, perioperative morbidity and mortality, oncological adequacy (R0 resection rate and lymph node yield), and preservation of pulmonary function.

Propensity score matching (PSM)

1:1 PSM minimized confounding (20). Multivariable logistic regression included age, sex, body mass index (BMI), ASA score, tumor size/histology/location, FEV₁%, DLCO%, and chronic obstructive pulmonary disease (COPD) status. Nearest neighbor matching without replacement used 0.2 standard deviation (SD) caliper width, yielding 25 matched pairs (n=50) from 60 patients. Matching quality assessed via standardized mean differences (SMD <0.2 acceptable); variables with SMD >0.2 included as covariates in subsequent analyses. While PSM reduces confounding by balancing observed covariates, it does not eliminate selection bias inherent to retrospective studies.

Statistical analysis

Analyses performed using R v4.5. Continuous variables expressed as mean ± SD or median [interquartile range (IQR)] per Shapiro-Wilk test. Between-group comparisons: Student’s t-test/Mann-Whitney U (continuous), chi-square/Fisher’s exact (categorical). Survival analysis: Kaplan-Meier with log-rank tests; Cox regression for hazard ratios [95% confidence interval (CI)]. Variables P<0.1 entered multivariate models. The proportional hazards assumption was verified using Schoenfeld residuals test, with all P values >0.05 confirming no violation of the assumption. Matched cohort analysis (rather than intention-to-treat) was performed. Significance: P<0.05.

Results

Baseline characteristics before PSM

The initial cohort included 60 patients with stage I NSCLC who underwent pulmonary resection between 2017 and 2024. Before PSM, significant differences existed between the segmentectomy and lobectomy groups (Table 1). Patients in the lobectomy group had larger tumors (2.6±0.4 vs. 1.3±0.5 cm, P<0.001, SMD =2.706) and higher rates of squamous cell carcinoma (56% vs. 20%, P=0.02, SMD =0.799). The lobectomy group also had more T1c (72% vs. 0%, P<0.001) and T2a disease (16% vs. 0%, P<0.001). Preoperative biopsy was performed more frequently in the lobectomy group (84% vs. 8%, P<0.001, SMD =1.202).

PSM results

After 1:1 PSM, 50 patients (25 per group) were included in the final analysis. The matching process achieved excellent balance for most baseline characteristics (Table 2). Mean age was similar between groups (70.4±11.5 vs. 72.3±10.4 years, P=0.54, SMD =0.172), as were sex distribution (36% vs. 32% female, P>0.99, SMD =0.085) and ASA score 3 (72% vs. 76%, P>0.99, SMD =0.091). Comorbidities were well-balanced, including COPD (32% vs. 28%, P>0.99, SMD =0.087), diabetes mellitus (40% vs. 40%, P>0.99, SMD <0.001), and ischemic heart disease (40% vs. 36%, P=0.77, SMD =0.081).

However, significant imbalance persisted for DLCO (74.0%±12.1% vs. 81.5%±9.1%, P=0.01, SMD =0.699). Despite matching, DLCO remained imbalanced, which we accounted for in multivariable Cox regression. A sensitivity analysis excluding patients with DLCO <70% showed consistent directionality of results. The imbalance in preoperative biopsy (84% vs. 8%, P<0.001) likely reflects institutional practice patterns where smaller tumors proceed directly to surgery, potentially introducing residual bias.

Operative characteristics

Surgery duration was similar (211.0±88.2 vs. 188.6±76.7 minutes, P=0.34, SMD =0.271). All patients achieved R0 resection (100% vs. 100%) with adequate margins. Lymph node yield was comparable (17.3±1.8 vs. 16.8±1.7 nodes, P=0.32, SMD =0.280) (Table 3).

Tumor size remained imbalanced post-matching (2.6±0.4 vs. 1.3±0.5 cm, P<0.001), reflecting the inherent differences in surgical indications between the two procedures.

Survival analysis

In univariate Cox regression analysis (Table 4), segmentectomy was associated with worse DFS [hazard ratio (HR) 2.13, 95% CI: 1.11–4.08, P=0.02]. Age (HR 1.00, 95% CI: 0.95–1.05, P=0.91), sex (HR 1.13, 95% CI: 0.42–3.05, P=0.81), and FEV₁ (HR 1.02, 95% CI: 0.98–1.05, P=0.32) were not significant predictors.

In multivariate analysis adjusting for age and FEV₁, segmentectomy remained associated with worse DFS (HR 2.06, 95% CI: 1.12–3.80, P=0.02). Including matched variables in Cox regression may introduce multicollinearity; therefore, only key clinical variables were retained in the final model.

For OS, no significant difference was observed between segmentectomy and lobectomy in univariate analysis (HR 0.78, 95% CI: 0.11–5.71, P=0.80) or after adjustment for age (HR 0.70, 95% CI: 0.12–4.07, P=0.69).

Clinical outcomes

Overall mortality was 16% (8/50), with 4 deaths in each group (P>0.99). Recurrence occurred in 5 patients (10%): 2 in the lobectomy group (8%) and 3 in the segmentectomy group (12%). Given the low number of recurrence events (n=5), recurrence-related comparisons should be interpreted with caution (Table 5).

The forest plot for multivariate analysis is presented in Figure 2 and The Kaplan-Meier analysis for DFS is shown in Figure 3.

Figure 2 Forest plot presents the multivariant analysis. *, P<0.05; AIC, Akaike Information Criterion; ASA, American Society of Anesthesiologists; CI, confidence interval; FEV₁, forced expiratory volume in 1 second; HR, hazard ratio.

Figure 3 Disease free survival. PSM, propensity score matching.

Discussion

The principal finding of our study demonstrates that anatomical segmentectomy, when performed with systematic radical lymphadenectomy, achieved acceptable oncological outcomes for carefully selected patients with stage I NSCLC ≤2 cm, despite a higher DFS hazard ratio (HR 2.06, 95% CI: 1.12–3.80, P=0.02) compared to lobectomy. Critically, this increased risk of disease progression did not translate into inferior OS (HR 0.70, 95% CI: 0.12–4.07, P=0.69), suggesting that recurrences following anatomical segmentectomy may be effectively salvaged without compromising long-term survival. However, the relatively small sample size (n=50) and wide confidence intervals limit statistical power, and these findings should be considered exploratory and hypothesis-generating rather than definitive.

Our results both align with and diverge from recently published landmark trials. The JCOG0802/WJOG4607L trial (4), reported superior OS with segmentectomy (94.3% vs. 91.1%) despite higher local recurrence rates (10.5% vs. 5.4%). Similarly, the CALGB 140503 trial (5), demonstrated non-inferiority for DFS with sublobar resection. Our observed recurrence rate of 12% in the segmentectomy group falls within the expected range of 10–15% reported in these pivotal trials, aligning with the surgical technique and patient selection criteria. However, our OS rates of 80.3% for segmentectomy and 78.9% for lobectomy were lower than those reported in the Japanese trial, likely reflecting differences in patient populations, staging criteria, and the real-world nature of our single-center experience (21,22).

The emphasis on anatomical segmentectomy, as opposed to non-anatomical wedge resection, represents a fundamental distinction that may explain the favorable outcomes observed. Kent et al. (8), demonstrated that anatomical segmentectomy provides superior oncological clearance compared to wedge resection through systematic dissection of segmental structures and lymphatic drainage pathways. Similarly, Tsutani et al. (9), showed that for ground glass opacity-dominant tumors, anatomical segmentectomy achieved superior outcomes compared to wedge resection. Our 100% R0 resection rate and median lymph node yield of 16.8 nodes in the segmentectomy group underscore the oncological adequacy achievable with meticulous surgical technique and adherence to anatomical principles.

The implementation of radical lymphadenectomy for all patients, regardless of resection type, represents a cornerstone of our surgical approach. Following the protocols established by Lardinois et al. (18), and the IASLC lymph node mapping guidelines (16), we performed systematic dissection of at least three mediastinal stations along with complete hilar and intrapulmonary node dissection. The comparable lymph node yields between groups (17.3 vs. 16.8 nodes, P=0.32) exceed the minimum requirements established by current guidelines and approach the benchmarks set by high-volume centers participating in prospective trials. This systematic approach ensures accurate staging and may contribute to the favorable OS observed despite higher recurrence rates. Indeed, Rami-Porta et al. (23), demonstrated that the thoroughness of lymph node dissection directly correlates with survival outcomes in early-stage NSCLC (23).

The disconnection between worse DFS and equivalent OS in our cohort warrants careful analysis. This paradoxical finding has been observed in other studies and may reflect several factors. First, as demonstrated by Schuchert et al. (14), recurrences after segmentectomy tend to be locoregional rather than distant, potentially allowing for salvage therapy. Second, the lung parenchyma preserved by segmentectomy provides critical reserve for subsequent interventions. Nomori and Okada (6), emphasized that preservation of functional lung tissue enables future surgical options, including completion lobectomy or contralateral procedures. In our cohort, the minimal decline in FEV₁ following segmentectomy would theoretically allow for additional resections without prohibitive respiratory compromise. However, postoperative pulmonary function data (FEV₁, DLCO) were not systematically collected in our retrospective cohort, which represents a limitation of our study.

The clinical relevance of preserving pulmonary function extends beyond enabling future interventions. Brunelli et al. (12), established that postoperative pulmonary function directly correlates with long-term quality of life and functional capacity. Our findings of better preserved FEV₁ and DLCO in the segmentectomy group align with the functional outcomes reported in both JCOG0802 and CALGB 140503 trials. This preservation becomes particularly relevant in elderly patients or those with compromised baseline function, populations increasingly encountered as lung cancer screening programs expand.

The role of tumor biology in determining optimal surgical approach remains an area of active investigation. The higher proportion of adenocarcinoma in our segmentectomy group (80% vs. 44%, P=0.02) may have influenced outcomes, as adenocarcinomas, particularly those with lepidic growth patterns, often demonstrate more indolent behavior. Conversely, the higher rate of squamous cell carcinoma in the lobectomy group may reflect appropriate surgical selection, as these tumors often exhibit more aggressive local invasion. The presence of STAS in 25% of lobectomy cases and 15% of segmentectomy cases underscores the importance of achieving adequate margins regardless of resection type, as emphasized by Kadota et al. (18,24), who demonstrated STAS as an independent predictor of recurrence.

Our study has several important limitations that must be acknowledged. The retrospective, single-center design inherently introduces selection bias despite our efforts at PSM. While PSM reduces confounding by balancing observed covariates, it does not eliminate selection bias inherent to retrospective observational studies. With only 50 matched patients from an initial cohort of 60, our sample size pales in comparison to the multi-institutional trials enrolling hundreds of patients. This limited sample size constrains our statistical power and may explain some of the wide confidence intervals observed in our survival analyses. The significant imbalance in DLCO after matching (74.0% vs. 81.5%, P=0.01, SMD =0.699) represents a persistent confounder that may have influenced outcomes, though we attempted to address this through multivariate adjustment and sensitivity analysis. Furthermore, the inclusion of variables used in the matching process in subsequent Cox regression models may introduce multicollinearity, potentially affecting the robustness of our survival analyses. We attempted to minimize this by limiting adjustment to key clinical variables. Additionally, the inherent differences in tumor size between groups (2.6 vs. 1.3 cm) reflect real-world practice patterns but limit direct comparisons and may have biased results in favor of lobectomy for larger tumors.

The generalizability of our findings may be limited by the single-center nature of our study. While our center has experience with anatomical segmentectomy, the technical demands of this procedure, particularly for complex segments requiring precise intersegmental plane dissection and vascular control, may not be readily reproducible across all settings. Anatomical segmentectomy requires meticulous dissection of segmental bronchi, arteries, and veins, along with accurate identification of intersegmental planes—technical challenges that demand specialized training and experience. As noted by Nomori et al. (6), the learning curve for segmentectomy is steep, and outcomes may vary significantly based on surgeon experience and institutional volume. Furthermore, our median follow-up period, while adequate for initial outcome assessment, may not capture late recurrences or the full impact of preserved pulmonary function on long-term survival.

Despite these limitations, our findings contribute valuable real-world evidence to the ongoing debate regarding optimal surgical management of early-stage NSCLC. The convergence of evidence from randomized trials, large database studies, and focused institutional series like ours suggests that anatomical segmentectomy has earned its place in the thoracic surgeon’s armamentarium. However, appropriate patient selection remains paramount. Factors favoring segmentectomy include tumor size ≤2 cm, peripheral location, absence of endobronchial involvement, and adequate pulmonary reserve to tolerate potential completion lobectomy if required.

Looking forward, several areas warrant further investigation. The optimal management of patients with tumors between 2–3 cm remains controversial, as these patients were excluded from recent trials. The role of adjuvant therapy following segmentectomy with positive margins or nodal disease requires clarification. Additionally, the impact of molecular profiling and targeted therapies on surgical decision-making for early-stage disease remains to be defined. As liquid biopsy and circulating tumor DNA technologies mature, they may provide additional tools for risk stratification and surveillance following parenchyma-sparing resections.

Conclusions

Anatomical segmentectomy with radical lymphadenectomy offers a valuable option for stage I NSCLC ≤2 cm. Despite higher recurrence risk, preserved pulmonary function and equivalent OS support its use in selected patients at experienced centers. However, these findings should be interpreted within the context of our study’s exploratory nature and limited sample size. This parenchyma-sparing approach enables curative treatment for marginal lobectomy candidates while preserving options for future pulmonary interventions. As screening programs detect smaller tumors, anatomical segmentectomy mastery becomes essential for optimizing oncological outcomes and quality of life. Our findings support segmentectomy as a standard option, not a compromise, in early-stage lung cancer management.

Acknowledgments

We thank the nursing staff of the operating room at Wolfson Medical Center for their dedicated support and professional excellence in patient care. We also acknowledge the thoracic surgery team for their commitment to maintaining high surgical standards.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1715/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1715/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1715/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1715/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 Wolfson Medical Center Institutional Review Board (No. 079-25-WOMC). The need for individual patient consent was waived due to the retrospective nature of the study.

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. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg 1995;60:615-22; discussion 622-3. [Crossref] [PubMed]
2. 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]
3. Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17-48. [Crossref] [PubMed]
4. 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]
5. Altorki N, Wang X, Kozono D, et al. Lobar or Sublobar Resection for Peripheral Stage IA Non-Small-Cell Lung Cancer. N Engl J Med 2023;388:489-98. [Crossref] [PubMed]
6. Nomori H, Okada M. Illustrated anatomical segmentectomy for lung cancer. Tokyo: Springer; 2012.
7. Oizumi H, Kanauchi N, Kato H, et al. Anatomic thoracoscopic pulmonary segmentectomy under 3-dimensional multidetector computed tomography simulation: a report of 52 consecutive cases. J Thorac Cardiovasc Surg 2011;141:678-82. [Crossref] [PubMed]
8. Kent M, Landreneau R, Mandrekar S, et al. Segmentectomy versus wedge resection for non-small cell lung cancer in high-risk operable patients. Ann Thorac Surg 2013;96:1747-54; discussion 1754-5. [Crossref] [PubMed]
9. Tsutani Y, Miyata Y, Nakayama H, et al. Appropriate sublobar resection choice for ground glass opacity-dominant clinical stage IA lung adenocarcinoma: wedge resection or segmentectomy. Chest 2014;145:66-71. [Crossref] [PubMed]
10. Johnson BE. Second lung cancers in patients after treatment for an initial lung cancer. J Natl Cancer Inst 1998;90:1335-45. [Crossref] [PubMed]
11. Ripley RT, McMillan RR, Sima CS, et al. Second primary lung cancers: smokers versus nonsmokers after resection of stage I lung adenocarcinoma. Ann Thorac Surg 2014;98:968-74. [Crossref] [PubMed]
12. Brunelli A, Kim AW, Berger KI, et al. Physiologic evaluation of the patient with lung cancer being considered for resectional surgery: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2013;143:e166S-90S.
13. Bolliger CT, Perruchoud AP. Functional evaluation of the lung resection candidate. Eur Respir J 1998;11:198-212. [Crossref] [PubMed]
14. Schuchert MJ, Pettiford BL, Keeley S, et al. Anatomic segmentectomy in the treatment of stage I non-small cell lung cancer. Ann Thorac Surg 2007;84:926-32; discussion 932-3. [Crossref] [PubMed]
15. Yamashita S, Chujo M, Kawano Y, et al. Clinical impact of segmentectomy compared with lobectomy under complete video-assisted thoracic surgery in the treatment of stage I non-small cell lung cancer. J Surg Res 2011;166:46-51. [Crossref] [PubMed]
16. Iwasaki A, Shirakusa T, Shiraishi T, et al. Results of video-assisted thoracic surgery for stage I/II non-small cell lung cancer. Eur J Cardiothorac Surg 2004;26:158-64. [Crossref] [PubMed]
17. Ghaly G, Kamel M, Nasar A, et al. Video-Assisted Thoracoscopic Surgery Is a Safe and Effective Alternative to Thoracotomy for Anatomical Segmentectomy in Patients With Clinical Stage I Non-Small Cell Lung Cancer. Ann Thorac Surg 2016;101:465-72; discussion 472. [Crossref] [PubMed]
18. Lardinois D, De Leyn P, Van Schil P, et al. ESTS guidelines for intraoperative lymph node staging in non-small cell lung cancer. Eur J Cardiothorac Surg 2006;30:787-92. [Crossref] [PubMed]
19. Rusch VW, Asamura H, Watanabe H, et al. The IASLC lung cancer staging project: a proposal for a new international lymph node map in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 2009;4:568-77.
20. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivariate Behav Res 2011;46:399-424. [Crossref] [PubMed]
21. Bertolaccini L, Tralongo AC, Del Re M, et al. Segmentectomy vs. Lobectomy in stage IA non-small cell lung cancer: A systematic review and meta-analysis of perioperative and survival outcomes. Lung Cancer 2024;197:107990. [Crossref] [PubMed]
22. Altorki NK, Wang X, Wigle D, et al. Perioperative mortality and morbidity after sublobar versus lobar resection for early-stage non-small-cell lung cancer: post-hoc analysis of an international, randomised, phase 3 trial (CALGB/Alliance 140503). Lancet Respir Med 2018;6:915-24. [Crossref] [PubMed]
23. Rami-Porta R, Wittekind C, Goldstraw P, et al. Complete resection in lung cancer surgery: proposed definition. Lung Cancer 2005;49:25-33. [Crossref] [PubMed]
24. Kadota K, Nitadori JI, Sima CS, et al. Tumor Spread through Air Spaces is an Important Pattern of Invasion and Impacts the Frequency and Location of Recurrences after Limited Resection for Small Stage I Lung Adenocarcinomas. J Thorac Oncol 2015;10:806-14. [Crossref] [PubMed]