Histological grading in lung cancer: one system for all or separate systems for each histological type?

The International Association for the Study of Lung Cancer (IASLC) Staging Project for Lung Cancer: Recommendation to Introduce Spread Through Air Spaces as a Histologic Descriptor in the Ninth Edition of the TNM Classification of Lung Cancer. Analysis of 4061 Pathologic Stage I NSCLC

INTRODUCTION

Spread through air spaces (STAS) consists of lung cancer tumor cells that are identified beyond the edge of the main tumor in the surrounding alveolar parenchyma. It has been reported by meta-analyses to be an independent prognostic factor in the major histologic types of lung cancer, but its role in lung cancer staging is not established.

METHODS

To assess the clinical importance of STAS in lung cancer staging, we evaluated 4061 surgically resected pathologic stage I R0 NSCLC collected from around the world in the International Association for the Study of Lung Cancer database. We focused on whether STAS could be a useful additional histologic descriptor to supplement the existing ones of visceral pleural invasion (VPI) and lymphovascular invasion (LVI).

RESULTS

STAS was found in 930 of 4061 of the pathologic stage I NSCLC (22.9%). Patients with tumors exhibiting STAS had a significantly worse recurrence-free and overall survival in both univariate and multivariable analyses involving cohorts consisting of all NSCLC, specific histologic types (adenocarcinoma and other NSCLC), and extent of resection (lobar and sublobar). Interestingly, STAS was independent of VPI in all of these analyses.

CONCLUSIONS

These data support our recommendation to include STAS as a histologic descriptor for the Ninth Edition of the TNM Classification of Lung Cancer. Hopefully, gathering these data in the coming years will facilitate a thorough analysis to better understand the relative impact of STAS, LVI, and VPI on lung cancer staging for the Tenth Edition TNM Stage Classification.

Multidirectional characterization of cellular composition and spatial architecture in human multiple primary lung cancers

Multiple primary lung cancers (MPLCs) pose diagnostic and therapeutic challenges in clinic. Here, we orchestrated the cellular and spatial architecture of MPLCs by combining single-cell RNA-sequencing and spatial transcriptomics. Notably, we identified a previously undescribed sub-population of epithelial cells termed as CLDN2⁺ alveolar type II (AT2) which was specifically enriched in MPLCs. This subtype was observed to possess a relatively stationary state, play a critical role in cellular communication, aggregate spatially in tumor tissues, and dominate the malignant histopathological patterns. The CLDN2 protein expression can help distinguish MPLCs from intrapulmonary metastasis and solitary lung cancer. Moreover, a cell surface receptor-TNFRSF18/GITR was highly expressed in T cells of MPLCs, suggesting TNFRSF18 as one potential immunotherapeutic target in MPLCs. Meanwhile, high inter-lesion heterogeneity was observed in MPLCs. These findings will provide insights into diagnostic biomarkers and therapeutic targets and advance our understanding of the cellular and spatial architecture of MPLCs.

The prognostic value of histopathology in invasive lung adenocarcinoma: a comparative review of the main proposed grading systems

INTRODUCTION

An accurate histological evaluation of invasive lung adenocarcinoma is essential for a correct clinical and pathological definition of the tumour. Different grading systems have been proposed to predict the prognosis of invasive lung adenocarcinoma.

AREAS COVERED

Invasive non mucinous lung adenocarcinoma is often morphologically heterogeneous, consisting of complex combinations of architectural patterns with different proportions. Several grading systems for non-mucinous lung adenocarcinoma have been proposed, being the main based on architectural differentiation and the predominant growth pattern. Herein we perform a thorough review of the literature using PubMed, Scopus and Web of Science and we highlight the peculiarities and the differences between the main grading systems and compare the data about their prognostic value. In addition, we carried out an evaluation of the proposed grading systems for less common histological variants of lung adenocarcinoma, such as fetal adenocarcinoma and invasive mucinous adenocarcinoma.

EXPERT OPINION

The current IASLC grading system, based on the combined score of predominant growth pattern plus high-grade histological pattern, shows the stronger prognostic significance than the previous grading systems in invasive non mucinous lung adenocarcinoma.

Expression patterns of platinum resistance-related genes in lung adenocarcinoma and related clinical value models

The purpose of this study was to explore platinum resistance-related biomarkers and mechanisms in lung adenocarcinoma. Through the analysis of gene expression data of lung adenocarcinoma patients and normal patients from The Cancer Genome Atlas, Gene Expression Omnibus database, and A database of genes related to platinum resistance, platinum resistance genes in lung adenocarcinoma and platinum resistance-related differentially expressed genes were obtained. After screening by a statistical significance threshold, a total of 252 genes were defined as platinum resistance genes with significant differential expression, of which 161 were up-regulated and 91 were down-regulated. The enrichment results of up-regulated gene Gene Ontology (GO) showed that TOP3 entries related to biological processes (BP) were double-strand break repair, DNA recombination, DNA replication, the down-regulated gene GO enriches the TOP3 items about biological processes (BP) as a response to lipopolysaccharide, muscle cell proliferation, response to molecule of bacterial origin. Gene Set Enrichment Analysis showed that the top three were e2f targets, g2m checkpoint, and rgf beta signaling. A prognostic model based on non-negative matrix factorization classification showed the characteristics of high- and low-risk groups. The prognostic model established by least absolute shrinkage and selection operator regression and risk factor analysis showed that genes such as HOXB7, NT5E, and KRT18 were positively correlated with risk score. By analyzing the differences in m6A regulatory factors between high- and low-risk groups, it was found that FTO, GPM6A, METTL3, and YTHDC2 were higher in the low-risk group, while HNRNPA2B1, HNRNPC, TGF2BP1, IGF2BP2, IGF2BP3, and RBM15B were higher in the high-risk group. Immune infiltration and drug sensitivity analysis also showed the gene characteristics of the platinum-resistant population in lung adenocarcinoma. ceRNA analysis showed that has-miR-374a-5p and RP6-24A23.7 were lower in the tumor expression group, and that the survival of the low expression group was worse than that of the high expression group. In conclusion, the results of this study show that platinum resistance-related differentially expressed genes in lung adenocarcinoma are mainly concentrated in biological processes such as DNA recombination and response to lipopolysaccharide. The validation set proved that the high-risk group of our prognostic model had poor survival. M6A regulatory factor analysis, immune infiltration, and drug sensitivity analysis all showed differences between high and low-risk groups. ceRNA analysis showed that has-miR-374a-5p and RP6-24A23.7 could be protective factors. Further exploration of the potential impact of these genes on the risk and prognosis of drug-resistant patients with lung adenocarcinoma would provide theoretical support for future research.

Perftoran® Inhibits Hypoxia-Associated Resistance in Lung Cancer Cells to Carboplatin

Perftoran^® (perfluorodecalin) is an oxygen carrier, and carboplatin is a common chemotherapy drug used worldwide for lung cancer treatment. Hypoxia is one of the factors that induce resistance of lung cancer cells to carboplatin. This study explored the role of Perftoran^®, as an oxygen carrier, in lowering the resistance of lung cancer cells to carboplatin through suppression of hypoxia pathway mediators. The effect of Perftoran^® on the resistance of human lung cancer A549 cells to carboplatin was investigated through the evaluation of cytotoxicity by MTT, cell death mode by dual DNA staining, DNA damage by comet assay, DNA platination (DNA/carboplatin adducts) by atomic absorption spectroscopy, hypoxia degree by pimonidazole, HIF-1α/HIF-2α concentrations by ELISA, expression of miRNAs (hypoxamiRs miR-210, miR-21, and miR-181a) by qRT-PCR, and the content of drug resistance transporter MRP-2 by immunocytochemical staining. Results indicated that compared to carboplatin, Perftoran^®/carboplatin decreased cell resistance to carboplatin by potentiating its cytotoxicity using only 45% of carboplatin IC₅₀ and inducing apoptosis. Perftoran^® induced DNA platination and DNA damage index in cells compared to carboplatin alone. Moreover, compared to treatment with carboplatin alone, co-treatment of cells with Perftoran^® and carboplatin inhibited cellular pimonidazole hypoxia adducts, diminished HIF-1α/HIF-2α concentrations, suppressed hypoxamiR expression, and decreased MRP-2. In conclusion, Perftoran^® inhibited resistance of lung cancer cells to carboplatin through the inhibition of both hypoxia pathway mediators and the drug resistance transporter MRP-2 and through the induction of DNA/carboplatin adduct formation.

Texture features of primary tumor on 18F-FDG PET images in non-small cell lung cancer: The relationship between imaging and histopathological parameters

PURPOSE

The aims of this study were to evaluate the relationships between textural features of the primary tumor on FDG PET images and clinical-histopathological parameters which are useful in predicting prognosis in newly diagnosed non-small cell lung cancer (NSCLC) patients.

METHODS

PET/CT images of ninety (90) patients with NSCLC prior to surgery were analyzed retrospectively. All patients had resectable tumors. From the images we acquired data related to metabolism (SUVmax, MTV, TLG) and texture features of primary tumors. Histopathological tumor types and subgroups, degree of Ki-67 expression and necrosis rates of the primary tumor, mediastinal lymph node (MLN) status and nodal stages were recorded.

RESULTS

Among the two histologic tumor types (adenocarcinoma and squamous cell carcinoma) significant differences were present regarding metabolic parameters, Ki-67 index with higher values and kurtosis with lower values in the latter group. Textural heterogeneity was found to be higher in poorly differentiated tumors compared to moderately differentiated tumors in patients with adenocarcinoma. While Ki-67 index had significant correlations with metabolic parameters and kurtosis, tumor necrosis rate was only significantly correlated with textural features. By univariate and multivariate analyses of the imaging and histopathological factors examined, only gradient variance was significant predictive factor for the presence of MLN metastasis.

CONCLUSIONS

Textural features had significant associations with histologic tumor types, degree of pathological differentiation, tumor proliferation and necrosis rates. Texture analysis has potential to differentiate tumor types and subtypes and to predict MLN metastasis in patients with NSCLC.

Pathogenesis and therapeutic strategy in platinum resistance lung cancer

Platinum compounds (cisplatin and carboplatin) represent the most active anticancer agents in clinical use both of lung cancer in mono-and combination therapies. However, platinum resistance limits its clinical application. It is necessary to understand the molecular mechanism of platinum resistance, identify predictive markers, and develop newer, more effective and less toxic agents to treat platinum resistance in lung cancer. Here, it summarizes the main molecular mechanisms associated with platinum resistance in lung cancer and the development of new approaches to tackle this clinically relevant problem. Moreover, it could lead to the development of more effective treatment for refractory lung cancer in future.

Monoamine Oxidase A Inhibits Lung Adenocarcinoma Cell Proliferation by Abrogating Aerobic Glycolysis

Lung adenocarcinoma (LUAD) accounts for ~30% of all lung cancers and is one of the causes of cancer-related death worldwide. As the role of monoamine oxidase A (MAOA) in LUAD remains unclear, in this study, we examine how MAOA affects LUAD cell proliferation. Analyses of both public data and our data reveal that the expression of MAOA is downregulated in LUAD compared with non-tumor tissue. In addition, the expression of MAOA in tumors correlates with clinicopathologic features, and the expression of MAOA serves as an independent biomarker in LUAD. In addition, the overexpression of MAOA inhibits LUAD cell proliferation by inducing G1 arrest in vitro. Further mechanistic studies show that MAOA abrogates aerobic glycolysis in LUAD cells by decreasing hexokinase 2 (HK2). Finally, the expression of HK2 shows a negative correlation with MAOA in LUAD, and high HK2 predicts poor clinical outcome. In conclusion, our findings indicate that MAOA functions as a tumor suppressor in LUAD. Our results indicate that the MAOA/HK2 axis could be potential targets in LUAD therapy.

Nongenetic Evolution Drives Lung Adenocarcinoma Spatial Heterogeneity and Progression

Cancer evolution determines molecular and morphologic intratumor heterogeneity and challenges the design of effective treatments. In lung adenocarcinoma, disease progression and prognosis are associated with the appearance of morphologically diverse tumor regions, termed histologic patterns. However, the link between molecular and histologic features remains elusive. Here, we generated multiomics and spatially resolved molecular profiles of histologic patterns from primary lung adenocarcinoma, which we integrated with molecular data from >2,000 patients. The transition from indolent to aggressive patterns was not driven by genetic alterations but by epigenetic and transcriptional reprogramming reshaping cancer cell identity. A signature quantifying this transition was an independent predictor of patient prognosis in multiple human cohorts. Within individual tumors, highly multiplexed protein spatial profiling revealed coexistence of immune desert, inflamed, and excluded regions, which matched histologic pattern composition. Our results provide a detailed molecular map of lung adenocarcinoma intratumor spatial heterogeneity, tracing nongenetic routes of cancer evolution. SIGNIFICANCE: Lung adenocarcinomas are classified based on histologic pattern prevalence. However, individual tumors exhibit multiple patterns with unknown molecular features. We characterized nongenetic mechanisms underlying intratumor patterns and molecular markers predicting patient prognosis. Intratumor patterns determined diverse immune microenvironments, warranting their study in the context of current immunotherapies.This article is highlighted in the In This Issue feature, p. 1307.

Deep learning in digital pathology image analysis: a survey

Deep learning (DL) has achieved state-of-the-art performance in many digital pathology analysis tasks. Traditional methods usually require hand-crafted domain-specific features, and DL methods can learn representations without manually designed features. In terms of feature extraction, DL approaches are less labor intensive compared with conventional machine learning methods. In this paper, we comprehensively summarize recent DL-based image analysis studies in histopathology, including different tasks (e.g., classification, semantic segmentation, detection, and instance segmentation) and various applications (e.g., stain normalization, cell/gland/region structure analysis). DL methods can provide consistent and accurate outcomes. DL is a promising tool to assist pathologists in clinical diagnosis.

Prediction of Non-small Cell Lung Cancer Histology by a Deep Ensemble of Convolutional and Bidirectional Recurrent Neural Network

Histology subtype prediction is a major task for grading non-small cell lung cancer (NSCLC) tumors. Invasive methods such as biopsy often lack in tumor sample, and as a result radiologists or oncologists find it difficult to detect proper histology of NSCLC tumors. The non-invasive methods such as machine learning may play a useful role to predict NSCLC histology by using medical image biomarkers. Few attempts have so far been made to predict NSCLC histology by considering all the major subtypes. The present study aimed to develop a more accurate deep learning model by clubbing convolutional and bidirectional recurrent neural networks. The NSCLC Radiogenomics dataset having 211 subjects was used in the study. Ten best models found during experimentation were averaged to form an ensemble. The model ensemble was executed with 10-fold repeated stratified cross-validation, and the results got were tested with metrics like accuracy, recall, precision, F1-score, Cohen's kappa, and ROC-AUC score. The accuracy of the ensemble model showed considerable improvement over the best model found with the single model. The proposed model may help significantly in the automated prognosis of NSCLC and other types of cancers.

Histological Grade: Analysis of Prognosis of Non-small Cell Lung Cancer After Complete Resection

BACKGROUND/AIM

Although the 2015 World Health Organization Classification reported that histological grading may be helpful in lung cancer management, a widely accepted histological grading system with clearly defined criteria and demonstrable clinical significance has not been developed. We investigated the prognoses of patients with resected non-small cell lung cancer (NSCLC) to identify prognostic factors, especially histological grade.

MATERIALS AND METHODS

The medical records of 531 patients between 2010 and 2015 were retrospectively reviewed. Overall survival (OS) curve was plotted using the Kaplan-Meier method. Cox regression analyses were used to evaluate the hazard ratio (HR) with endpoint of OS.

RESULTS

The 5-year OS rate in groups with histological grade 1, grade 2, and grade 3+4 groups was 95.8%, 85.7%, and 72.1%, respectively (p<0.001). Multivariate analysis identified histological grade and vascular invasion as independent predictors of OS [histological grade: HR=1.533, p=0.002].

CONCLUSION

Histological grade was an independent prognostic factor of patients resected for all stages of NSCLC.

Programmed death ligand 1 expression in early stage, resectable non-small cell lung cancer

INTRODUCTION

For several years non-small cell lung cancer (NSCLC) has been considered non-immunogenic. Recent advances in antitumor immunity brought to the discovery of checkpoints that modulate immune response against cancer. One of them is programmed death receptor 1 (PD-1) and its ligand (PD-L1). Although PD-L1 expression seems predictive of response to anti-PD-1/PD-L1 agents, its prognostic value is unclear. In this study we investigated the prognostic value of PD-L1 expression and its correlation with clinical-pathological characteristics in a cohort of surgically resected NSCLC.

MATERIAL AND METHODS

PD-L1 expression was evaluated in 289 surgically resected NSCLC samples by immunohistochemistry. Our cohort included patients not exposed to adjuvant chemotherapy. PD-L1 status was defined as: 1) PD-L1 high (tumor proportion score, TPS≥50%), PD-L1 low (TPS 1-49%), PD-L1 negative (TPS<1%); 2) PD-L1 positive (TPS≥50%) and negative (TPS<50%); 3) as a continuous variable.

RESULTS

Patients were mostly males (79%), former or current smokers (81%), with a median age of 67 years, non-squamous histology (67.5%) and high-grade tumors (55%). PD-L1 tumors were 18.7%. There was no significant association with sex, age, smoking status and histology. A strong correlation between high PD-L1 expression and tumor grade was detected. The difference in median OS in the different groups of patients was not statistically significant.

CONCLUSION

PD-L1 is not prognostic in surgically resected NSCLC. The association with tumor differentiation suggests that grading could represent an easy-to-assess tool for selecting subjects potentially sensitive to immunotherapy warranting further investigations.

A Direct Plasma miRNA Assay for Early Detection and Histological Classification of Lung Cancer

Cell-free microRNAs in plasma provide circulating biomarkers for lung cancer. Most techniques for analysis of miRNAs require a large plasma volume to purify a sufficient RNA yield followed by complicated downstream processing. Small differences in the multiple procedures often cause large analytical variations and poor diagnostic values of the plasma biomarkers. Here we investigate whether directly quantifying plasma miRNAs without RNA purification could diagnose lung cancer. FirePlex assay was directly applied to 20 μl plasma of 56 lung cancer patients and 28 cancer free controls for quantifying 11 lung tumor–associated miRNAs. FirePlex assay is easier, less expensive and time-consuming for quantification of plasma miRNAs compared with conventional reverse transcription PCR with an equivalent analytic performance. From the lung tumor–associated miRNAs, a prediction model based on two miRNAs (miRs-205-5p and -210-3p) was developed, producing 78.6% sensitivity and 89.3% specificity for identifying lung cancer. The diagnostic value was independent of stage of lung tumor, and patients’ age and sex (all P > 0.05). Furthermore, based on the same two miRNAs, additional prediction models were developed with 75.0% sensitivity and 89.3% specificity for diagnosis of lung squamous cell carcinoma, and 82.2% sensitivity and 89.3% specificity for lung adenocarcinoma. The direct plasma assay can improve the efficacy of miRNA assessment in a small plasma volume by reducing multiple procedure-associated analytical variables. The developed plasma miRNA biomarkers might be useful for the early detection and histological classification of lung cancer.

An overview of mortality & predictors of small-cell and non-small cell lung cancer among Saudi patients

Lung cancer ranks as the top cancer worldwide in terms of incidence and constitutes a major health problem. About 90% of lung cancer cases are diagnosed at advance stage where treatment is not available. Despite evidence that lung cancer screening improves survival, guidelines for lung cancer screening are still a subject for debate. In Saudi Arabia, only 14% of lung cancers are diagnosed at early stage and researches on survival and its predictors are lacking. This overview analysis was conducted on predictors of lung cancer mortality according to the two major cancer types, small-cell lung cancers (SCLCs) and non-small cell lung cancers (NSCLCs) in Saudi Arabia. A secondary data analysis was performed on small-cell lung cancers (SCLCs) and Non-small cell lung cancers (NSCLCs) registered in the Saudi Cancer Registry (SCR) for the period 2009-2013 to estimate predictors of mortality for both lung cancer types. A total of 404 cases (197 SCLC and 207 NSCLC) were included in the analysis, all Saudi nationals. A total of 213 (52.75%) deaths occurred among lung cancer patients, 108 (54.82%) among SCLCs and 105 (50.72%) among NCSLCs. Three quarter of patients are diagnosis with advance stage for both SCLC & NSCLC. Univariate analysis revealed higher mean age at diagnosis in dead patients compared to alive patients for SCLCs (p=0.04); but not NSCLCs, a lower mortality for NSCLCs diagnosed in 2013 (p=0.025) and a significant difference in stage of tumor (p=0.006) and (p=0.035) for both SCLC and NSCLC respectively. In multiple logistic regression, stage of tumor was a strong predictor of mortality, where distant metastasis increased morality by 6-fold (OR=5.87, 95% CI: 2.01 - 17.19) in SCLC and by 3-fold (OR=3.29, 95% CI: 1.22 - 8.85) in NSCLC, compared to localized tumors. Those with NSCLC who were diagnosed in 2013 were less likely to die by 64% compared to NSCLC diagnosed in 2009 (OR=0.36, 95% CI: 0.14 - 0.93). Age, sex, topography and laterality were not associated with mortality for both types of lung cancer. We observed that the stage of the tumor is the strongest predictor of mortality for both SCLCs and NSCLs. This confirms the impact of diagnostic stage on survival. However, establishing Saudi-specific lung cancer screening guidelines will require further research on the benefits and harms of screening modalities in the Saudi population.

Analysis of the proliferative activity in lung adenocarcinomas with specific driver mutations

In the last decade it became evident that many lung adenocarcinomas (ADC) harbor key genetic alterations such as KRAS, EGFR or BRAF mutations as well as rearrangements of ROS1 or ALK that drive these tumors. In the present study we investigated whether different driver mutations of ADC result in different proliferation rates, which might have clinical impact, including resistance to therapy, recurrence and prognosis. We analyzed the proliferation index (PI) on full slides of surgically resected ADC (n = 230) with known genetic aberrations by means of immunohistochemistry and subsequent digital image analysis and correlated the results with clinicopathological variables including overall (OS) and disease free survival (DFS). We did not observe significant differences in OS or DFS regarding the KRAS or EGFR mutational status (P = 0.56). However, KRAS mutated ADC showed an increased PI compared to EGFR mutated ADC, and ADC with ALK translocations (P < 0.01). Subgroup analysis of EGFR mutated ADC showed a higher PI for tumors harboring a mutation in exon 18 and 20, compared to tumors with a mutation in exon 19 or 21. A PI of 11.5% was the best possible prognostic stratificator for OS (P = 0.01 in KRAS mutated and P < 0.01 in EGFR mutated ADC). In conclusion, the PI differs significantly among ADC with distinct driver mutations. This might explain the varying indications for a prognostic relevance of the PI observed in prior studies. Our study provides a basis for the establishment of a reliable and clinically meaningful PI threshold.

[Grading of lung cancer]

In comparison with other tumor entities there is no common generally accepted grading system for lung cancer with clearly defined criteria and clinical relevance. In the recent fourth edition of the World Health Organization (WHO) classification from 2015 of tumors of the lungs, pleura, thymus and heart, there is no generally applicable grading for pulmonary adenocarcinomas, squamous cell carcinomas or rarer forms of carcinoma. Since the new IASLC/ATS/ERS classification of adenocarcinomas published in 2011, 5 different subtypes with significantly different prognosis are proposed. This results in an architectural (histologic) grading, which is usually applied to resection specimens. For squamous cell carcinoma the number of different histological subtypes in the new WHO classification was reduced compared to earlier versions but without a common grading system. In recent publications nesting and budding were proposed as the main (histologic) criteria for a grading of squamous cell carcinomas. The grading of neuroendocrine tumors (NET) of the lungs in comparison with NET in other organs is presented in a separate article in this issue. Certain rare tumor types are high grade per definition: small cell, large cell and pleomorphic carcinomas, carcinosarcomas and pulmonary blastomas. In the future it is to be expected that these developments will be further refined, e. g. by adding further subtypes for adenocarcinomas and cytologic and/or nuclear criteria for adenocarcinoma and/or squamous cell carcinomas.

A prediction model for distinguishing lung squamous cell carcinoma from adenocarcinoma

Accurate classification of squamous cell carcinoma (SCC) from adenocarcinoma (AC) of non–small cell lung cancer (NSCLC) can lead to personalized treatments of lung cancer. We aimed to develop a miRNA-based prediction model for differentiating SCC from AC in surgical resected tissues and bronchoalveolar lavage (BAL) samples. Expression levels of seven histological subtype-associated miRNAs were determined in 128 snap-frozen surgical lung tumor specimens by using reverse transcription-polymerase chain reaction (RT-PCR) to develop an optimal panel of miRNAs for acutely distinguishing SCC from AC. The biomarkers were validated in an independent cohort of 112 FFPE lung tumor tissues, and a cohort of 127 BAL specimens by using droplet digital PCR for differentiating SCC from AC. A prediction model with two miRNAs (miRs-205-5p and 944) was developed that had 0.988 area under the curve (AUC) with 96.55% sensitivity and 96.43% specificity for differentiating SCC from AC in frozen tissues, and 0.997 AUC with 96.43% sensitivity and 96.43% specificity in FFPE specimens. The diagnostic performance of the prediction model was reproducibly validated in BAL specimens for distinguishing SCC from AC with a higher accuracy compared with cytology (95.69 vs. 68.10%; P < 0.05). The prediction model might have a clinical value for accurately discriminating SCC from AC in both surgical lung tumor tissues and liquid cytological specimens.