AI-driven Characterization of Solid Pulmonary Nodules on CT Imaging for Enhanced Malignancy Prediction in Small-sized Lung Adenocarcinoma.

OBJECTIVES: Distinguishing solid nodules from nodules with ground-glass lesions in lung cancer is a critical diagnostic challenge, especially for tumors ≤2 cm. Human assessment of these nodules is associated with high inter-observer variability, which is why an objective and reliable diagnostic tool is necessary. This study focuses on artificial intelligence (AI) to automatically analyze such tumors and to develop prospective AI systems that can independently differentiate highly malignant nodules. MATERIALS AND METHODS: Our retrospective study analyzed 246 patients who were diagnosed with negative clinical lymph node metastases (cN0) using positron emission tomography-computed tomography (PET/CT) imaging and underwent surgical resection for lung adenocarcinoma. AI detected tumor sizes ≤2 cm in these patients. By utilizing AI to classify these nodules as solid (AI_solid) or non-solid (non-AI_solid) based on confidence scores, we aim to correlate AI determinations with pathological findings, thereby advancing the precision of preoperative assessments. RESULTS: Solid nodules identified by AI with a confidence score ≥0.87 showed significantly higher solid component volumes and proportions in patients with AI_solid than in those with non-AI_solid, with no differences in overall diameter or total volume of the tumors. Among patients with AI_solid, 16% demonstrated lymph node metastasis, and a significant 94% harbored invasive adenocarcinoma. Additionally, 44% were upstaging postoperatively. These AI_solid nodules represented high-grade malignancies. CONCLUSION: In small-sized lung cancer diagnosed as cN0, AI automatically identifies tumors as solid nodules ≤2 cm and evaluates their malignancy preoperatively. The AI classification can inform lymph node assessment necessity in sublobar resections, reflecting metastatic potential.

Long-term prognosis analysis of surgical therapy for bilateral synchronous multiple primary lung cancer: a follow-up of 293 cases.

Background: Bilateral synchronous multiple primary lung cancer (BSMPLC) presents significant clinical challenges due to its unique characteristics and prognosis. Understanding the risk factors that influence overall survival (OS) and recurrence-free survival (RFS) is crucial for optimizing therapeutic strategies for BSMPLC patients. Methods: We retrospectively analyzed clinical characteristics and treatment outcomes of 293 patients with BSMPLC who underwent surgical treatment between January 2010 and July 2017. Results: The 10-year OS and RFS rates were 96.1% and 92.8%, respectively. Preoperative forced expiratory volume in 1 second (FEV1) ≥70% [hazard ratio (HR), 0.214; 95% confidence interval (CI): 0.053 to 0.857], identical pathology types (HR, 9.726; 95% CI: 1.886 to 50.151), largest pT1 (HR, 7.123; 95% CI: 2.663 to 19.055), and absence of lymphovascular invasion (LVI; HR, 7.021; 95% CI: 1.448 to 34.032) emerged as independent predictors of improved OS. Moreover, the sum of tumor sizes less than or equal to 3 cm (HR, 6.229; 95% CI: 1.411 to 27.502) and absence of pleural invasion (HR, 3.442; 95% CI: 1.352 to 8.759) were identified as independent predictors of enhanced RFS. The presence or absence of residual nodules after bilateral surgery did not influence patients' OS (P=0.987) and RFS (P=0.054). Conclusions: Patients with BSMPLC who underwent surgery generally had a favorable prognosis. Whether or not to remove all nodules bilaterally does not affect the patient's long-term prognosis, suggesting the need for an individualized surgical approach.

Sublobar Resection in Early Non-Small Cell Lung Cancer With Epidermal Growth Factor Receptor Mutant.

BACKGROUND: Lobectomy is a standard surgical procedure for peripherally located early-stage non-small cell lung cancers (NSCLCs) measuring 2 to 4 cm. However, it is unclear whether sublobar resections, such as wedge resection and segmentectomy, are effective in treating tumors with driver mutations in the epidermal growth factor receptor (EGFR). METHODS: We analyzed the clinicopathologic findings and surgical outcomes of 1395 patients with radiologically solid-dominant NSCLC measuring 2 to 4 cm, without clinical lymph node involvement, who underwent complete resection between 2010 and 2020. The patients, who underwent sublobar resections (n = 231) or lobectomy (n = 1164), were categorized by their EGFR mutation status and the surgical procedures performed. The follow-up was conducted for a median of 45.3 months. RESULTS: The 5-year overall survival (OS) rates after sublobar resections (n = 39) were comparable to those after lobectomy (n = 359) in patients with EGFR mutation-positive tumors (80.5% [95% CI, 51.3%-93.2%] vs 88.8% [95% CI, 84.1%-92.1%], respectively; P = .16). Multivariable Cox regression analysis of OS revealed that the surgical procedure was an independent prognostic predictor in the entire cohort (hazard ratio, 0.6; 95% CI, 0.4-1.0; P = .028), but it was not an independent prognostic predictor in patients with EGFR-mutated tumors (hazard ratio, 0.6; 95% CI, 0.2-1.7; P = .32). CONCLUSIONS: Sublobar resection with a secure surgical margin could be a viable option for appropriately selected patients with peripheral early-stage NSCLC tumors measuring 2 to 4 cm and harboring EGFR mutations, because it provides comparable OS to that of lobectomy.

Risk Factors for Recurrence of Stage I Epidermal Growth Factor Receptor Mutated Lung Adenocarcinoma.

BACKGROUND: We aimed to clarify the risk factors for postoperative recurrence in patients with epidermal growth factor receptor (EGFR)-mutated stage I lung adenocarcinoma, using EGFR wild-type adenocarcinoma as a comparator, to select optimal candidates for adjuvant therapy with EGFR tyrosine kinase inhibitor (TKI). METHODS: Data of patients with pathologic stage I EGFR-mutated (n = 713) and wild-type (n = 673) adenocarcinoma who did not receive adjuvant therapy were retrospectively analyzed. The cumulative incidence of recurrence (CIR) was estimated using Gray's method, and multivariable Fine-Gray competing risk models identified independent risk factors associated with recurrence. RESULTS: The CIR did not differ significantly between patients with EGFR-mutated and wild-type adenocarcinoma (P = .32). Multivariable analysis revealed that greater size (cm) of invasive tumor (hazard ratio 1.539; 95% CI, 1.077-2.201), lymphovascular invasion (hazard ratio 5.180; 95% CI, 2.208-12.15), pleural invasion (hazard ratio 3.388; 95% CI, 1.524-7.533), and high-grade histologic subtype (hazard ratio 4.295; 95% CI, 1.539-11.99) were independent risk factors for recurrence in patients with EGFR-mutated adenocarcinoma. The 5-year CIR was significantly higher among patients with these factors (tumor size greater than 2 cm, 15.9%; lymphovascular invasion, 26.9%; pleural invasion, 39.3%; and high-grade subtype, 44.4%) than among patients without them (4.4%, 2.2%, 3.9%, and 5%, respectively; P < .001). For patients with EGFR wild-type adenocarcinoma, independent risk factors for recurrence were invasive tumor size, lymphovascular invasion, and pleural invasion, but not histologic subtypes. CONCLUSIONS: Even for patients with EGFR-mutated stage I lung adenocarcinoma, recurrence risk is stratified. Adjuvant therapy may be considered if they have high-risk factors for recurrence.

A multicenter propensity score-matched analysis of lymphadenectomy in N1-positve lung cancer.

OBJECTIVES: Selective mediastinal lymph node dissection based on lobe-specific metastases is widely recognized in daily practice. However, the significance of mediastinal lymph node dissection for N1-positive tumors has not been elucidated. METHODS: We retrospectively reviewed 359 patients with N1-positive lung cancer who underwent lobectomy with systematic mediastinal lymph node dissection (systematic lymph node dissection) (n = 150) and lobe-specific mediastinal lymph node dissection (lobe-specific lymph node dissection) (n = 209). The operative and postoperative results and their propensity score-matched pairs were compared. The factors affecting survival were assessed using competing risk and multivariable analyses. RESULTS: The cumulative incidence of recurrence and the cumulative incidence of cancer-specific death were not significantly different between systematic and lobe-specific lymph node dissection in entire cohort. In the propensity score-matched cohort (83 pairs), systematic lymph node dissection tended to detect N2 lymph node metastasis more frequently (55.4 vs. 41%, P = 0.087). Eleven patients (13.2%) in the systematic lymph node dissection group had a metastatic N2 lymph node 'in the systematic lymph node dissection field' that lobe-specific lymph node dissection did not dissect. The oncological outcomes between patients undergoing systematic lymph node dissection (5-year cumulative incidence of recurrence, 62.1%; 5-year cumulative incidence of cancer-specific death, 27.9%) and lobe-specific lymph node dissection (5-year cumulative incidence of recurrence, 60.1%; 5-year cumulative incidence of cancer-specific death, 23.3%) were similar. The propensity score-adjusted multivariable analysis for cumulative incidence of recurrence revealed that the prognosis associated with systematic lymph node dissection was comparable with the prognosis with lobe-specific lymph node dissection (hazard ratio, 1.17; 95% confidence interval, 0.82-1.67; P = 0.37). CONCLUSIONS: The extent of lymph node dissection can affect accurate pathological staging; however, it was not associated with survival outcome in the treatment of N1-positive lung cancer.

Segmentectomy for cancer control in radiologically pure-solid clinical stage IA3 lung cancer.

OBJECTIVES: This study aimed to compare cancer control after segmentectomy and lobectomy in patients with radiologically pure-solid clinical stage IA3 non-small-cell lung cancer (NSCLC). METHODS: Patients with radiologically pure-solid clinical stage IA3 NSCLC who underwent lobectomy or segmentectomy at 3 institutions between 2010 and 2019 were identified. We estimated propensity scores to adjust for confounding variables regarding tumour malignancy, including age, sex, smoking history, tumour size, maximum standardized uptake value on 18F-fluorodeoxyglucose positron emission tomography, lymph node dissection, histological type and lymphatic, vascular and pleural invasion. Cumulative incidence of recurrence (CIR) was evaluated as a primary end point. RESULTS: Among 412 patients, postoperative recurrence occurred in 7 of 44 patients (15.9%) undergoing segmentectomy, and 71 of 368 patients (19.3%) undergoing lobectomy. CIR was comparable between patients undergoing segmentectomy (5-year rate, 21.9%) and those undergoing lobectomy (5-year rate, 20.8%; P = 0.88). Locoregional recurrence did not differ between patients undergoing segmentectomy (6.8%) and those undergoing lobectomy (9.0%). In multivariable analysis, segmentectomy (versus lobectomy) was not identified as an independent prognostic factor for CIR (hazard ratio, 1.045; 95% confidence interval, 0.475-2.298; P = 0.91). In propensity score matching of 40 pairs, CIR was not significantly different between patients undergoing segmentectomy (5-year rate, 20.7%) and those undergoing lobectomy (5-year rate, 18.4%; P = 0.81). CONCLUSIONS: Cancer control may be comparable between segmentectomy and lobectomy in patients with radiologically pure-solid clinical stage IA3 NSCLC. Further studies are warranted to clarify the survival benefits of segmentectomy in these patients.

Extracellular vesicle-associated microRNA signatures related to lymphovascular invasion in early-stage lung adenocarcinoma.

Lymphovascular invasion (LVI) is a fundamental step toward the spread of cancer. Extracellular vesicles (EVs) promote cellular communication by shuttling cargo, such as microRNAs (miRNAs). However, whether EV-associated miRNAs serve as biomarkers for LVI remains unclear. This study aimed to identify EV-associated miRNAs related to LVI and validate the miRNA levels from patients with early-stage lung adenocarcinoma (LADC). Blood samples were collected from patients undergoing pulmonary resection for stage I LADC before surgery. The patients were classified into three groups according to the presence of LVI and postoperative recurrence. Serum-derived EVs in the derivation cohort were used for small RNA sequencing, while the selected LVI miRNA candidates were validated via real-time quantitative polymerase chain reaction using 44 patient and 16 healthy donor samples as the validation cohorts. Five miRNAs (miR-99b-3p, miR-26a-5p, miR-93-5p, miR-30d-5p, and miR-365b-3p) were assessed, and miR-30d-5p (p = 0.036) levels were significantly downregulated in the LVI-positive group. miR-30d-5p levels in healthy donors were lower than those in LADC patients. Patients with high miR-30d-5p levels had favorable survival compared to those with low miR-30d-5p levels. miR-30d-5p level in EVs may serve as a promising biomarker for detecting LVI in patients with early-stage LADC.

Wedge resection is an acceptable treatment option for radiologically low-grade lung cancer with solid predominance.

OBJECTIVES: This study aimed to determine the clinical characteristics for predicting low-grade cancer in radiologically solid predominant non-small-cell lung cancer (NSCLC) and compare the survival outcomes of wedge resection with those of anatomical resection for patients with and without these characteristics. METHODS: Consecutive patients with clinical stages IA1-IA2 NSCLC showing radiologically solid predominance ≤2 cm at 3 institutions were retrospectively evaluated. Low-grade cancer was defined as the absence of nodal involvement and blood vessel, lymphatic and pleural invasion. The predictive criteria for low-grade cancer were established by multivariable analysis. The prognosis of wedge resection was compared with that of anatomical resection for patients who met the criteria, using the propensity score-matched analysis. RESULTS: Among 669 patients, multivariable analysis showed that ground-glass opacity (GGO) (P < 0.001) on thin-section computed tomography and an increased maximum standardized uptake value on 18-fluoro-2-deoxyglucose positron emission tomography/computed tomography (P < 0.001) were independent predictors of low-grade cancer. The predictive criteria were defined as GGO presence and maximum standardized uptake value ≤1.1 (specificity: 97.8%, sensitivity: 21.4%). In the propensity score-matched pairs (n = 189), overall survival (P = 0.41) and relapse-free survival (P = 0.18) were not significantly different between patients who underwent wedge resection and anatomical resection among those who fulfilled the criteria. CONCLUSIONS: The radiologic criteria for GGO and a low maximum standardized uptake value could predict low-grade cancer, even in solid-dominant NSCLC sized ≤2 cm. Wedge resection could be an acceptable surgical option for patients with radiologically predicted indolent NSCLC showing a solid-dominant appearance.

Intraplaque wiring enables drug-coated balloons to be utilized for percutaneous recanalization of chronically occluded coronary arteries.

OBJECTIVES: This study aimed to determine whether drug-coated balloon (DCB) angioplasty following intraplaque wiring and the use of modified balloons is safe and effective in the percutaneous treatment of coronary chronic total occlusions (CTOs). BACKGROUND: DCB is an alternative therapeutic option without the limitations of permanent vascular implants. However, its efficacy in CTOs has yet to be confirmed. The combination of modified balloons and DCB can be effectively applied when the intraplaque passage of the guidewire is achieved in CTOs. METHODS: Data from 124 consecutive CTO lesions (105 patients) treated at our hospital between February 2016 and December 2020 were screened for inclusion and retrospectively analyzed. Among the 118 lesions successfully recanalized, intraplaque wiring was achieved in 108, and 85 were treated by the DCB-only approach following cutting/scoring balloon dilatation. RESULTS: Follow-up data were available for 82 lesions (71 patients). The median occlusion length was 18.5 mm, and the J-CTO score was 1.7 ± 0.9. No in-hospital major adverse cardiac events occurred, including abrupt vessel closure. During the median 29-month follow-up period, target lesion revascularization was performed for 10 lesions. Follow-up coronary angiography (8.7 ± 3.9 months after the index procedure) was performed for 64 lesions, demonstrating late lumen loss of -0.15 mm (interquartile range -0.4 to 0.23 mm), binary restenosis (diameter stenosis ≥50%) in 12 lesions (18.8%), and late lumen enlargement in 37 (57.8%). CONCLUSION: The DCB-only approach following the use of modified balloons is a promising strategy for coronary CTOs when intraplaque wiring is achieved.

Artificial intelligence-based radiomics for the prediction of nodal metastasis in early-stage lung cancer.

We aimed to investigate the value of computed tomography (CT)-based radiomics with artificial intelligence (AI) in predicting pathological lymph node metastasis (pN) in patients with clinical stage 0-IA non-small cell lung cancer (c-stage 0-IA NSCLC). This study enrolled 720 patients who underwent complete surgical resection for c-stage 0-IA NSCLC, and were assigned to the derivation and validation cohorts. Using the AI software Beta Version (Fujifilm Corporation, Japan), 39 AI imaging factors, including 17 factors from the AI ground-glass nodule analysis and 22 radiomics features from nodule characterization analysis, were extracted to identify factors associated with pN. Multivariate analysis showed that clinical stage IA3 (p = 0.028), solid-part size (p < 0.001), and average solid CT value (p = 0.033) were independently associated with pN. The receiver operating characteristic analysis showed that the area under the curve and optimal cut-off values of the average solid CT value relevant to pN were 0.761 and -103 Hounsfield units, and the threshold provided sensitivity, specificity, and negative predictive values of 69%, 65%, and 94% in the entire cohort, respectively. Measuring the average solid-CT value of tumors for pN may have broad applications such as guiding individualized surgical approaches and postoperative treatment.

Clinical utility of psoas muscle volume in assessment of sarcopenia in patients with early-stage non-small cell lung cancer.

PURPOSE: Sarcopenia influences postoperative outcomes of patients with non-small cell lung cancer (NSCLC). Imaging tools for evaluating and diagnosing sarcopenia have developed, and a novel method of psoas volume index (PVI) obtained by measuring bilateral psoas major muscle volume has been reported. However, the relationship between sarcopenia based on PVI and clinical outcomes has not been fully investigated for patients with early-stage NSCLC. This study aimed to clarify the utility of PVI values in assessing the relationshipe between sarcopenia and clinical outcomes. METHODS: This study included 645 patients with stage I-II NSCLC who underwent curative lung resection between 2012 and 2017. Bilateral psoas major muscle volumes were calculated semi-automatically using a three-dimensional workstation. The cutoff value of PVI for defining sarcopenia was < 60.5 cm3/m3 for men and < 43.6 cm3/m3 for women. RESULTS: The avrage time to obtaine PVI was only 25 s with the 3D system, and interobserver agreements for evauating sarcopenia on PVI was 1. A total of 159 patients (24.7%) were preoperatively diagnosed with sarcopenia. On multivariate analysis, sarcopenia was an independent prognostic factor for overall survival (OS, p < 0.001), recurrence-free survival (RFS, p < 0.001), and lung cancer-specific survival (LCS, p < 0.001). The 5-year OS, RFS, and LCS were significantly worse in sarcopenic patients than non-sarcopenic patients (88.8 vs. 72.4%, p < 0.001; 80.1 vs. 65.0%, p < 0.001; 92.4 vs. 78.9%, p < 0.001, respectively). CONCLUSION: Sarcopenia diagnosed using PVI is an independent prognostic predictor of OS, RFS, and LCS in early-stage NSCLC.

Discrepancy Between Radiological and Pathological Tumor Size in Early-Stage Non-Small Cell Lung Cancer: A Multicenter Study.

Discrepancies between radiological whole tumor size (RTS) and pathological whole tumor size (PTS) are sometimes observed. Unexpected pathological upsize may lead to insufficient margins during procedures like sub lobar resections. Therefore, this study aimed to investigate the current status of these discrepancies and identify factors resulting in pathological upsize in patients with early-stage non-small cell lung cancer (NSCLC). Data from a multicenter database of 3092 patients with clinical stage 0-IA NSCLC who underwent pulmonary resection were retrospectively analyzed. Differences between the RTS and PTS were evaluated using Pearson's correlation analysis and Bland-Altman plots. Unexpected pathological upsize was defined as an upsize of ≥1 cm when compared to the RTS, and the predictive factors of this upsize were identified based on multivariable analyses. The RTS and PTS showed a positive linear relationship (r = 0.659), and the RTS slightly overestimated the PTS. The Bland-Altman plot showed 131 of 3092 (5.2%) cases were over the upper 95% limits of agreement. In multivariable analyses, a maximum standardized uptake value (SUVmax) of the primary tumor on 18-fluoro-2-deoxyglucose positron emission tomography/computed tomography (odds ratio [OR], 1.070; 95% confidence interval [CI], 1.035-1.107; P < 0.001) and the adenocarcinoma histology (OR, 1.899; 95% CI, 1.071-3.369; P =0.049) were independent predictors of unexpected pathological upsize. More of the adenocarcinomas with pathological upsize were moderately or poorly differentiated, when compared to those without. The RTS tends to overestimate the PTS; however, care needs to be taken regarding unexpected pathological upsize, especially in adenocarcinomas with a high SUVmax.

Expert consensus on indocyanine green fluorescence imaging for thoracoscopic lung resection (The Version 2022).

The use of the white-light thoracoscopy is hampered by the low contrast between oncologic margins and surrounding normal parenchyma. As a result, many patients with in situ or micro-infiltrating adenocarcinoma have to undergo lobectomy due to a lack of tactile and visual feedback in the resection of solitary pulmonary nodules. Near-infrared (NIR) guided indocyanine green (ICG) fluorescence imaging technique has been widely investigated due to its unique capability in addressing the current challenges; however, there is no special consensus on the evidence and recommendations for its preoperative and intraoperative applications. This manuscript will describe the development process of a consensus on ICG fluorescence-guided thoracoscopic resection of pulmonary lesions and make recommendations that can be applied in a greater number of centers. Specifically, an expert panel of thoracic surgeons and radiographers was formed. Based on the quality of evidence and strength of recommendations, the consensus was developed in conjunction with the Chinese Guidelines on Video-assisted Thoracoscopy, and the National Comprehensive Cancer Network (NCCN) guidelines on the management of pulmonary lesions. Each of the statements was discussed and agreed upon with a unanimous consensus amongst the panel. A total of 6 consensus statements were developed. Fluorescence-guided thoracoscopy has unique advantages in the visualization of pulmonary nodules, and recognition and resection of the anterior plane of the pulmonary segment. The expert panel agrees that fluorescence-guided thoracoscopic surgery has the potential to become a routine operation for the treatment of pulmonary lesions.

Segmentectomy for clinically early-stage primary squamous cell carcinoma of the lung.

BACKGROUND: Squamous cell carcinoma of the lung-the second most common subtype of lung cancer-has a poorer prognosis than lung adenocarcinoma. However, in contrast to lobectomy, the oncological outcomes after segmentectomy for primary squamous cell carcinomas remain unknown; hence, this study investigated these outcomes. METHODS: Patients who underwent lobectomy or segmentectomy for clinically node-negative primary lung squamous cell carcinoma with a whole tumor size of ≤ 30 mm on preoperative computed tomography scan during April 2010 to December 2020 were included in this study. The cumulative incidence of recurrence (CIR) among all included patients and propensity score-matched patients were compared using the Gray method. Multivariate analysis using propensity scores and surgical procedures was performed using the Fine and Gray method. RESULTS: Overall, 230 patients were included in this study; of these, 172 (74.8%) underwent lobectomy and 58 (25.2%) underwent segmentectomy. No significant differences were observed in the CIR between patients who underwent lobectomy and those who underwent segmentectomy (5-year rate 18.1% vs. 14.2%; p  = â€Š0.787). Moreover, no significant differences in CIR were observed between the propensity score-matched patients who underwent lobectomy (n = 43) and those who underwent segmentectomy (n   = â€Š43) (8.6% vs. 8.0%; p = 0.571). Multivariable analysis was performed for CIR using the propensity score; it revealed that segmentectomy was not a significant predictor of worse CIR (hazard ratio, 0.987; p = â€Š 0.980). CONCLUSIONS: Segmentectomy may be feasible for treating clinically early-stage lung squamous cell carcinoma; its oncological outcomes are similar to those of lobectomy.

Radiomics with Artificial Intelligence for the Prediction of Early Recurrence in Patients with Clinical Stage IA Lung Cancer.

BACKGROUND: We seek to explore the ability of computed tomography (CT)-based radiomics coupled with artificial intelligence (AI) to predict early recurrence (< 2 years after surgery) in patients with clinical stage 0-IA non-small cell lung cancer (c-stage 0-IA NSCLC). PATIENTS AND METHODS: Data of 642 patients were collected for early recurrence and assigned to the derivation and validation cohorts at a ratio of 2:1. Using the AI software Beta Version (Fujifilm Corporation, Japan), 39 AI imaging factors, including 17 factors from the AI ground-glass nodule analysis and 22 radiomic features from nodule characterization analysis, were extracted. RESULTS: Multivariate analysis showed that male sex (p = 0.016), solid part size (p < 0.001), CT value standard deviation (p = 0.038), solid part volume ratio (p = 0.016), and bronchus translucency (p = 0.007) were associated with recurrence-free survival (RFS). Receiver operating characteristics analysis showed that the area under the curve and optimal cutoff values relevant to recurrence were 0.707 and 1.49 cm for solid part size, and 0.710 and 22.9% for solid part volume ratio, respectively. The 5-year RFS rates for patients in the validation set with solid part size ≤ 1.49 cm and > 1.49 cm were 92.2% and 70.4% (p < 0.001), whereas those for patients with solid part volume ratios ≤ 22.9% and > 22.9% were 97.8% and 71.7% (p < 0.001), respectively. CONCLUSIONS: CT-based radiomics coupled with AI contributes to the noninvasive prediction of early recurrence in patients with c-stage 0-IA NSCLC.