Objectives To investigate the clinicoradiological and pathological-molecular characteristics associated with spread through air spaces (STAS) in stage IA invasive lung adenocarcinoma (ILADC). Materials and methods We retrospectively analyzed clinicoradiological data from 765 surgically resected stage IA ILADC patients. STAS-positive/-negative group comparisons identified independent predictors for model development, which were externally validated. Model-derived threshold stratified patients into risk groups for RFS analysis. Additionally, pathological and molecular tumor characteristics were compared between STAS-positive and STAS-negative cases. Results Univariate analysis revealed significant differences in three clinical characteristics and eight radiological features between STAS-positive and STAS-negative cases (p < 0.05). Binary logistic regression identified CTR ≥ 90%, bronchial truncation, pleural retraction, and satellite nodules as independent predictors of STAS. The model yielded area under the curve (AUC) 0.823/accuracy 76.70% (training cohort) and AUC 0.829/accuracy 70.10% (external validation cohort). High-risk patients, as classified by the model, had significantly shorter RFS in the training and external validation cohorts (all p < 0.05). Additionally, STAS-positive tumors had a higher prevalence of micropapillary/solid-predominant growth and KRAS/ALK mutations, while EGFR mutations were more common in STAS-negative tumors (p < 0.05). Conclusion The proposed model demonstrates potential in preoperatively identifying STAS and stratifying recurrence risk for patients with stage IA ILADC, which may assist in selecting appropriate candidates for sublobar resection. Furthermore, the distinct pathological and molecular characteristics of STAS-positive tumors offer valuable insights for guiding personalized adjuvant therapy. Critical relevance statement This validated clinicoradiological model utilizes preoperative CT to predict STAS and recurrence risk, optimizing surgical decision-making. Furthermore, it explores these underlying pathological-molecular features associated with STAS, supporting personalized adjuvant therapy selection. Key Points <UnorderedList Mark="Bullet"> <ItemContent> Preoperative identification of STAS remains a clinical challenge in stage IA lung adenocarcinoma. </ItemContent> <ItemContent> Our clinicoradiological model can simultaneously predict STAS and stratify recurrence risk. </ItemContent> <ItemContent> STAS-positive cases can be characterized by micropapillary/solid patterns and KRAS/ALK mutations. </ItemContent> <ItemContent> These findings may assist with surgical decision-making and facilitate tailored adjuvant therapy selection. </ItemContent> </UnorderedList> Graphical Abstract

The clinicoradiological and pathological-molecular characteristics associated with spread through air spaces in stage IA invasive lung adenocarcinoma

期刊论文
出版模式：

开放获取
发表日期：

2026年6月24日

查看全文

Feng Xu,
Xian Li,
Wei-hua Zhao,
Ji-wen Huo,
Tian-you Luo,
Qi Li

摘要

Objectives

To investigate the clinicoradiological and pathological-molecular characteristics associated with spread through air spaces (STAS) in stage IA invasive lung adenocarcinoma (ILADC).

Materials and methods

We retrospectively analyzed clinicoradiological data from 765 surgically resected stage IA ILADC patients. STAS-positive/-negative group comparisons identified independent predictors for model development, which were externally validated. Model-derived threshold stratified patients into risk groups for RFS analysis. Additionally, pathological and molecular tumor characteristics were compared between STAS-positive and STAS-negative cases.

Results

Univariate analysis revealed significant differences in three clinical characteristics and eight radiological features between STAS-positive and STAS-negative cases (p < 0.05). Binary logistic regression identified CTR ≥ 90%, bronchial truncation, pleural retraction, and satellite nodules as independent predictors of STAS. The model yielded area under the curve (AUC) 0.823/accuracy 76.70% (training cohort) and AUC 0.829/accuracy 70.10% (external validation cohort). High-risk patients, as classified by the model, had significantly shorter RFS in the training and external validation cohorts (all p < 0.05). Additionally, STAS-positive tumors had a higher prevalence of micropapillary/solid-predominant growth and KRAS/ALK mutations, while EGFR mutations were more common in STAS-negative tumors (p < 0.05).

Conclusion

The proposed model demonstrates potential in preoperatively identifying STAS and stratifying recurrence risk for patients with stage IA ILADC, which may assist in selecting appropriate candidates for sublobar resection. Furthermore, the distinct pathological and molecular characteristics of STAS-positive tumors offer valuable insights for guiding personalized adjuvant therapy.

Critical relevance statement

This validated clinicoradiological model utilizes preoperative CT to predict STAS and recurrence risk, optimizing surgical decision-making. Furthermore, it explores these underlying pathological-molecular features associated with STAS, supporting personalized adjuvant therapy selection.

Key Points

Preoperative identification of STAS remains a clinical challenge in stage IA lung adenocarcinoma.

Our clinicoradiological model can simultaneously predict STAS and stratify recurrence risk.

STAS-positive cases can be characterized by micropapillary/solid patterns and KRAS/ALK mutations.

These findings may assist with surgical decision-making and facilitate tailored adjuvant therapy selection.

Graphical Abstract