A novel image deep learning-based sub-centimeter pulmonary nodule management algorithm to expedite resection of the malignant and avoid over-diagnosis of the benign.

OBJECTIVES: With the popularization of chest computed tomography (CT) screening, there are more sub-centimeter (≤ 1 cm) pulmonary nodules (SCPNs) requiring further diagnostic workup. This area represents an important opportunity to optimize the SCPN management algorithm avoiding "one-size fits all" approach. One critical problem is how to learn the discriminative multi-view characteristics and the unique context of each SCPN. METHODS: Here, we propose a multi-view coupled self-attention module (MVCS) to capture the global spatial context of the CT image through modeling the association order of space and dimension. Compared with existing self-attention methods, MVCS uses less memory consumption and computational complexity, unearths dimension correlations that previous methods have not found, and is easy to integrate with other frameworks. RESULTS: In total, a public dataset LUNA16 from LIDC-IDRI, 1319 SCPNs from 1069 patients presenting to a major referral center, and 160 SCPNs from 137 patients from three other major centers were analyzed to pre-train, train, and validate the model. Experimental results showed that performance outperforms the state-of-the-art models in terms of accuracy and stability and is comparable to that of human experts in classifying precancerous lesions and invasive adenocarcinoma. We also provide a fusion MVCS network (MVCSN) by combining the CT image with the clinical characteristics and radiographic features of patients. CONCLUSION: This tool may ultimately aid in expediting resection of the malignant SCPNs and avoid over-diagnosis of the benign ones, resulting in improved management outcomes. CLINICAL RELEVANCE STATEMENT: In the diagnosis of sub-centimeter lung adenocarcinoma, fusion MVCSN can help doctors improve work efficiency and guide their treatment decisions to a certain extent. KEY POINTS: • Advances in computed tomography (CT) not only increase the number of nodules detected, but also the nodules that are identified are smaller, such as sub-centimeter pulmonary nodules (SCPNs). • We propose a multi-view coupled self-attention module (MVCS), which could model spatial and dimensional correlations sequentially for learning global spatial contexts, which is better than other attention mechanisms. • MVCS uses fewer huge memory consumption and computational complexity than the existing self-attention methods when dealing with 3D medical image data. Additionally, it reaches promising accuracy for SCPNs' malignancy evaluation and has lower training cost than other models.

Spatial downregulation of CD74 signatures may drive invasive component development in part-solid lung adenocarcinoma.

Pulmonary nodules with part-solid imaging features manifest during the progression from preinvasive to invasive lung adenocarcinoma. To define the spatial composition and evolutionary trajectories of early-stage lung adenocarcinoma, we combined spatial transcriptomics (ST) and pathological annotations from 20 part-solid nodules (PSNs), four of which were matched with single-cell RNA sequencing. Two malignant cell populations (MC1 and MC2) were identified, and a linear evolutionary relationship was observed. Compared to MC2, the pre-existing malignant MC1 exhibited a lower metastatic signature, corresponding to the preinvasive component (lepidic) on pathology and the ground glass component on PSN imaging. Higher immune infiltration was observed among MC1 regions in ST profiles, and further analysis revealed that macrophages may be involved in this process through the CD74 axis. This work provides deeper insights into the evolutionary process and spatial immune cell composition behind PSNs and highlights the mechanisms of immune escape behind this adenocarcinoma trajectory.

The regularity of anatomical variations of dominant pulmonary segments in the right upper lobe.

BACKGROUND: Anatomical variations often pose challenges to pulmonary surgery. Previous studies have mainly described the frequencies of bronchovascular anatomical variations in pulmonary segments, but did not determine the differences between pulmonary segments and the regularity behind these anatomical variations. Here, we attempted to investigate the regularity of bronchovascular anatomical variations in different pulmonary segments. METHODS: Thin-slice enhanced computed tomography data of 800 cases from our center were included in this study. Digitalized three-dimensional virtual lung segmentation was done, the dominant and inferior lung segments of the right upper lobe were defined, and the regularity of anatomical variations was explored. RESULTS: The mean volume ratio of the anterior segment of the right upper lobe (39.6 ± 8.6%) was highest, and that of the posterior segment (28.6 ± 7.9%) was lowest. Therefore, the dominant-type segment (DS + SDS) was dominant in the anterior segment, accounting for 74.6% (597/800), and the inferior-type segment (SIS + IS) was dominant in the posterior segment of the right upper lobe, accounting for 71.5% of cases (573/800). During the transformation of dominant and inferior lung segments, the corresponding regularity of anatomical variations could be displayed. For example, with an increase in the volume of the anterior segment of the right upper lobe, the occurrence rate of the bifurcated type of bronchus (B1 + 2, B3), the "central vein type" and the involvement of the trunk inferior and ascending artery in the blood supply of anterior segment gradually increased. CONCLUSIONS: The existence of dominant segments will increase the diversity of anatomical variations and the complexity of pulmonary segmentectomy.

Sublobectomy for stage IA1-2 invasive lung adenocarcinoma with consolidation tumor ratio ≤ 0.25.

BACKGROUND: Sublobectomy for early-stage non-small cell lung cancer (NSCLC) remains a matter of debate. This study aimed to discuss the feasibility of sublobectomy in patients with pathological-stage IA1-2 confirmed as pathologically invasive but radiologically noninvasive adenocarcinoma. METHODS: From 2011 to 2019, we screened clinical stage IA1-IA2 lung cancer patients who underwent surgery at the Guangdong Provincial People's Hospital (GDPH). Inclusion criteria were maximum tumor diameter of 2.0 cm or less, consolidation tumor ratio (CTR) ≤ 0.25, and pathologically confirmed invasive adenocarcinoma. Sublobectomy (segmentectomy and wedge resection) and lobectomy groups were created, and propensity scores were computed. The primary endpoints were lung cancer-specific overall survival (LCSS) and LCS- relapse-free survival (LCS-RFS) after adjusting propensity scores. RESULTS: A total of 1731 patients were screened, and 100 patients were enrolled. The lobectomy group had 51 patients and the limited resection group had 49. No cases relapsed, and two patients died from nontumor causes. For the entire cohort, the 5-year LCSS and 5-year LCS-RFS were 100% in the lobectomy and limited resection groups. When propensity scores matched, there were no differences in LCSS and LCS-RFS between the two groups (LCSS:100%, LCS-RFS 100% in lobectomy and limited resection, respectively). DISCUSSION: Sublobectomy may be curative for pathologically invasive but radiologically noninvasive adenocarcinoma at pathological stage IA1-2.

Investigation on the incidence and risk factors of lung cancer among Chinese hospital employees.

OBJECTIVE: In recent years, the lung cancer incidence has grown and the population is younger. We intend to find out the true detection rate of pulmonary nodules and the incidence of lung cancer in the population and search for the risk factors. METHOD: Hospital employees ≥40 years old who underwent low-dose computed tomography (CT) lung cancer screening from January 2019 to March 2022 were selected to record CT-imaging characteristics, pathology, staging, and questionnaires to investigate past history, smoking history, diet, mental health, etc. PM2.5 and radiation intake in radiation-related occupation received monitoring in hospital. RESULT: The detection rate of suspicious pulmonary nodules was 9.1% (233/2552), and the incidence rate of lung cancer (including adenocarcinoma in situ) was 4.0% (103/2552). Morbidity among doctors, nurses, technicians, administers, and logistics was no difference (p = 0.184), but higher in women than in men (4.7% vs 2.4% p = 0.002). The invasiveness increased with age and CT density of nodules (p = 0.018). The relationship between lung cancer morbidity and PM2.5 was not clear (p = 0.543); and no lung cancer has been found in employees related ionizing radiation. CONCLUSION: The high screening rate has brought about a high incidence of lung cancer. At present, the risk factor analysis of lung cancer based on small samples cannot find the direct cause. Most of the ground glass opacity (GGO)s detected by LDCT screening are indolent, but there are also rapidly progressive lung cancer. A predictive model to identify active and indolent GGO is necessary.

Identification of heritable rare variants associated with early-stage lung adenocarcinoma risk.

Background: In East Asia, the number of patients with adenocarcinoma, especially those presenting with ground-glass nodules (GGNs), is gradually increasing. Family aggregation of pulmonary GGNs is not uncommon; however, genetic predisposition in these patients remains poorly understood and identification of genes involved in the cause of these early-stage lung cancers might contribute to understanding of the underlying mechanisms and potential prevention strategies. Methods: Fifty patients with early-stage lung adenocarcinoma (LUAD) presenting as GGNs and a first-degree family history of lung cancer (FHLC) from 34 independent families were enrolled into this study. Germline mutations of these patients were analyzed with whole exome sequencing (WES) and compared with age- and sex-matched 39 patients with sporadic lung cancer and 689 local healthy people. We used a stepwise variant filtering strategy, gene-based burden testing, and enrichment analysis to investigate rare but potentially pathogenic heritable mutations. Somatic tumor mutations were analyzed to consolidate germline findings. Results: In total, 1,571 single nucleotide variants (SNVs) and 238 frameshifts with a minor allele frequency (MAF) <0.01, which were rare, recurrent, and potentially damaging candidates, were finally identified through the filtering in the GGN cohort. Pathway analysis showed the extracellular matrix to be the top dysregulated pathway. Gene-based burden testing of these highly disruptive risk-conferring heritable variants showed that MSH5 [odds ratio (OR), 9.28, 95% confidence interval (CI): 2.49-35.87], MMP9 (OR, 8.11, 95% CI: 2.22-28.43), and CYP2D6 (OR, 8.09, 95% CI: 2.68-24.92) were significantly enriched in our cohort (P<0.05). The number of rare damaging germline variants in non-smoking patients was significantly higher than that of smoking-affected patients (Spearman's ρ=-0.39, P=0.02). Conclusions: Heritable, potentially deleterious, and rare candidate variants of MSH5, MMP9 and CYP2D6 were significantly associated with early-stage LUAD presenting with GGNs. Nonsmoking patients likely have a higher genetic predisposition to this type of cancer than smoking-affected patients. These results have extended our understanding of the underlying mechanisms of early-stage LUAD.

Clinicopathological features and resistance mechanisms in HIP1-ALK-rearranged lung cancer: A multicenter study.

Anaplastic lymphoma kinase (ALK)-rearranged non-small cell lung cancer (NSCLC) respond well to ALK tyrosine kinase inhibitors (TKIs), and echinoderm microtubule-associated protein-like 4 (EML4)-ALK-rearranged NSCLC accounts for the majority of those patients. However, few studies have evaluated ALK-TKIs treatment for patients with huntingtin-interacting protein 1 (HIP1)-ALK fusions. This retrospective study evaluated the clinicopathological characteristics, genomic features, response to ALK-TKIs, and resistance mechanisms in 11 cases with HIP1-ALK fusions from five Chinese centers. Patients who received crizotinib at the Chinese centers had an objective response rate of 90% [9/10 cases, 95% confident index (CI): 54.1%-99.5%], median progression-free survival of 17.9 months (95% CI: 5.8-NA months), and median overall survival of 58.8 months (95% CI: 24.7-NA months). One patient who received first-line lorlatinib treatment achieved partial response for > 26.5 months. Despite the small sample size, HIP1-ALK (H21:A20) variant was the most common variant (four of 11 cases, 36.4%) and associated with better outcomes. Among the 11 cases, there were eight patients having available specimens for genetic testing before ALK-TKIs treatment and four patients undergoing biopsy after ALK-TKIs failure. The most common coexisting gene was TP53 among 11 patients and two of four patients after crizotinib failure harbored acquired ALK mutations (e.g., L1152V/Q1146K and L1196M). Brigatinib treatment appeared to be effective for a patient who failed crizotinib treatment because of the L1152V/Q1146K mutations, which might be related to increased binding affinity to these mutants. Although HIP1-ALK-rearranged NSCLC appears to initially respond well to ALK-TKIs, crizotinib resistance may be correlated with the AKAP9-BRAF fusion, ALK compound mutations (L1152V/Q1146K), and the ALK L1196M mutation. Larger studies are needed to evaluate the significance of HIP1-ALK-rearranged NSCLC.

Quantifying invasiveness of clinical stage IA lung adenocarcinoma with computed tomography texture features.

OBJECTIVES: The study objectives were to establish and validate a nomogram for pathological invasiveness prediction in clinical stage IA lung adenocarcinoma and to help identify those potentially unsuitable for sublobar resection-based computed tomography texture features. METHOD: Patients with clinical stage IA lung adenocarcinoma who underwent surgery at Guangdong Provincial People's Hospital between January 2015 and October 2018 were retrospectively reviewed. All surgically resected nodules were pathologically classified into less-invasive and invasive cohorts. Each nodule was manually segmented, and its computerized texture features were extracted. Clinicopathological and computed tomographic texture features were compared between 2 cohorts. A nomogram for distinguishing the pathological invasiveness was established and validated. RESULTS: Among 428 enrolled patients, 249 were diagnosed with invasive pathological subtypes. Smoking status (odds ratio, 2.906; 95% confidence interval, 1.285-6.579; P = .011), mean computed tomography attenuation value (odds ratio, 1.005, 95% confidence interval, 1.002-1.007; P < .001), and entropy (odds ratio, 8.536, 95% confidence interval, 3.478-20.951; P < .001) were identified as independent predictors for pathological invasiveness by multivariate logistics regression analysis. The nomogram showed good calibration (P = .182) with an area under the curve of 0.849 when validated with testing set data. Decision curve analysis indicated the potentially clinical usefulness of the model with respect to treat-all or treat-none scenario. Compared with intraoperative frozen-section, the nomogram performed better in pathological invasiveness diagnosis (area under the curve, 0.815 vs 0.670; P = .00095). CONCLUSIONS: We established and validated a nomogram to compute the probability of invasiveness of clinical stage IA lung adenocarcinoma with great calibration, which may contribute to decisions related to resection extent.

Watershed analysis of the target pulmonary artery for real-time localization of non-palpable pulmonary nodules.

BACKGROUND: Some pulmonary nodules are not suitable for computed tomography-guided percutaneous localization. This study aimed to investigate the feasibility and safety of real-time localization for these non-palpable pulmonary nodules using watershed analysis of the target pulmonary artery during thoracoscopic wedge resection. METHODS: Watershed analysis is a novel technique that can be used to create a specific area on the lung surface for nodule localization. This analysis is performed by temporarily blocking the target pulmonary artery and using indocyanine green fluorescence during surgery. In our study, the surgery was simulated and evaluated preoperatively using a high-precision three-dimensional reconstruction model obtained by multidetector spiral computed tomography. The lung was observed using an infrared thoracoscopy system after an intravenous injection of indocyanine green (2.5 mg/mL), and the white-to-blue transitional zone was marked using electrocautery, after which a wedge resection was performed. RESULTS: A total of 25 out of 26 patients underwent successful wedge resection. The mean tumor size and depth based on computed tomography scans were 13.2±6.4 and 12.2±7.8 mm, respectively. The mean operation duration was 142.6±52.8 min. The mean bleeding volume during surgery was 12.9±9.7 mL. The mean drainage tube indwelling time was 35.6±20.0 h, and the median length of postoperative stay was 3 days (range, 2-6 days). CONCLUSIONS: Our experience showed that the watershed analysis of the target pulmonary artery for nodule localization was safe and feasible. It may become an effective and attractive alternative method for localizing non-palpable pulmonary nodules in selected patients undergoing thoracoscopic wedge resection.

Timing and Origins of Local and Distant Metastases in Lung Cancer.

INTRODUCTION: Metastasis is the primary cause of lung cancer-related death. Nevertheless, the underlying molecular mechanisms and evolutionary patterns of lung cancer metastases are still elusive. METHODS: We performed whole-exome sequencing for 40 primary tumors (PTs) and 61 metastases from 47 patients with lung cancer, of which 40 patients had paired PTs and metastases. The PT-metastasis genomic divergence, metastatic drivers, timing of metastatic dissemination, and evolutionary origins were analyzed using appropriate statistical tools and mathematical models. RESULTS: There were various degrees of genomic heterogeneity when comparing the paired primary and metastatic lesions or comparing metastases of different sites. Multiple metastasis-selected/enriched genetic alterations were found, such as MYC amplification, NKX2-1 amplification, RICTOR amplification, arm 20p gain, and arm 11p loss, and these results were were also featured in a meta-analysis cross-validated using an independent cohort from Memorial Sloan-Kettering Cancer Center database. To elucidate the metastatic seeding time, we applied a metastatic model and found 61.1% of the tumors were late dissemination, in which the metastatic seeding happened approximately 2.74 years before clinical detection. One exception was lymph node metastases whose dissemination time was relatively early. By analyzing the evolutionary origins, we reported that nonlymph node metastases were mainly seeded by the PT (87.5%) rather than the earlier colonized lymph node metastases. CONCLUSIONS: Our results shed light on the molecular features that potentially drive lung cancer metastases. The distinct temporospatial pattern of disease progression revealed that lung cancer was susceptible to either late dissemination or indolent early lymph node metastases, leaving a potential time window to minimize metastases by early cancer detection.

A three-dimensional printing navigational template combined with mixed reality technique for localizing pulmonary nodules.

OBJECTIVES: Localizing non-palpable pulmonary nodules is challenging for thoracic surgeons. Here, we investigated the accuracy of three-dimensional (3D) printing technology combined with mixed reality (MR) for localizing ground glass opacity-dominant pulmonary nodules. METHODS: In this single-arm study, we prospectively enrolled patients with small pulmonary nodules (<2 cm) that required accurate localization. A 3D-printing physical navigational template was designed based on the reconstruction of computed tomography images, and a 3D model was generated through the MR glasses. We set the deviation distance as the primary end point for efficacy evaluation. Clinicopathological and surgical data were obtained for further analysis. RESULTS: Sixteen patients with 17 non-palpable pulmonary nodules were enrolled in this study. Sixteen nodules were localized successfully (16/17; 94.1%) using this novel approach with a median deviation of 9 mm. The mean time required for localization was 25 ± 5.2 min. For the nodules in the upper/middle and lower lobes, the median deviation was 6 mm (range, 0-12.0) and 16 mm (range, 15.0-20.0), respectively. The deviation difference between the groups was significant (Z = -2.957, P = 0.003). The pathological evaluation of resection margins was negative. CONCLUSIONS: The 3D printing navigational template combined with MR can be a feasible approach for localizing pulmonary nodules.

Precise resection of multiple pulmonary nodules using a three-dimensional reconstruction model: A case report.

A 48-year-old woman presented to our department and chest computed tomography (CT) revealed five pulmonary nodules, two of which were in the left upper lobe of the lung and three in the superior segment of the left lower lobe., All the lesions were resected for comprehensive histological assessment in order to distinguish synchronous multiple primary lung cancers (SMPLCs) from intrapulmonary metastases. The nodules were all successfully removed by minimally invasive surgery under the guidance of three dimensional (3D) reconstruction, in order to preserve as much lung function for the patient as possible. Postoperative histopathological examination demonstrated the presence of SMPLC. The patient was discharged from hospital on postoperative day 4 without any complications.

Genomic characteristics and drug screening among organoids derived from non-small cell lung cancer patients.

BACKGROUND: Patient-derived organoid (PDO) models are highly valuable and have potentially widespread clinical applications. However, limited information is available regarding organoid models of non-small cell lung cancer (NSCLC). This study aimed to characterize the consistency between primary tumors in NSCLC and PDOs and to explore the applications of PDOs as preclinical models to understand and predict treatment response during lung cancer. METHODS: Fresh tumor samples were harvested for organoid culture. Primary tumor samples and PDOs were analyzed via whole-exome sequencing. Paired samples were subjected to immunohistochemical analysis. There were 26 antineoplastic drugs tested in the PDOs. Cell viability was assessed using the Cell Titer Glo assay 7-10 days after drug treatment. A heatmap of log-transformed values of the half-maximal inhibitory concentrations was generated on the basis of drug responses of PDOs through nonlinear regression (curve fit). A total of 12 patients (stages I-III) were enrolled, and 7 paired surgical tumors and PDOs were analyzed. RESULTS: PDOs retained the histological and genetic characteristics of the primary tumors. The concordance between tumors and PDOs in mutations in the top 20 NSCLC-related genes was >80% in five patients. Sample purity was significantly and positively associated with variant allele frequency (Pearson r = 0.82, P = 0.0005) and chromosome stability. The in vitro response to drug screening with PDOs revealed high correlation with the mutation profiles in the primary tumors. CONCLUSIONS: PDOs are highly credible models for detecting NSCLC and for prospective prediction of the treatment response for personalized precision medicine. KEY POINTS: Lung cancer organoid models could save precious time of drug testing on patients, and accurately select anticancer drugs according to the drug sensitivity results, so as to provide a powerful supplement and verification for the gene sequencing.

Randomized Trial of an Improved Drainage Strategy Versus Routine Chest Tube After Lung Wedge Resection.

BACKGROUND: Patients undergoing thoracic lung wedge resection could benefit from tubeless strategies. However, postoperative pneumothorax is a primary limiting factor for such strategies. Accordingly, we evaluated the safety and efficacy of the prophylactic use of an air-extraction catheter as an improved drainage strategy and compared the findings with those for chest tube drainage in patients undergoing thoracic wedge resection. METHODS: Patients undergoing thoracic wedge resection between August 2017 and October 2018 were enrolled in this single-center, randomized, open-label, noninferiority trial. Patients who received an improved drainage strategy involving the use of a prophylactic air-extraction catheter were randomized to the intervention group, whereas those who underwent routine chest tube drainage were assigned to the control group. Analysis was based on the per-protocol population. The primary outcome was the incidence of pneumothorax on postoperative day 1. Secondary outcomes included patient recovery and related complications, including pleural effusion, lung infection, numeric rating scale score for pain, postoperative chest tube or catheter removal, postoperative hospitalization, and chest tube reinsertion. RESULTS: A total of 96 patients were randomized. Baseline demographic and clinical characteristics were similar between groups. The incidence of pneumothorax in the intervention and control groups was 10.0% and 9.1%, respectively (noninferiority, P = 1.00). In addition, there were no significant between-group differences in secondary outcomes. A significantly lower pain score was observed in the intervention group (P = .001). CONCLUSIONS: The improved drainage strategy is not inferior to standard chest tube drainage after thoracic wedge resection and should be popularized.