Subtype Of Non-small Cell Lung Cancer Research Articles

Abstract 4888: Multi-modal knowledge graphs enhance patient stratification & biomarker discovery

Abstract Background: Multi-modal analysis is crucial for deeper understanding of disease subtypes and more meaningful patient selection. We developed a flexible Knowledge Graph (KG) framework that enables deep multi-omic analysis. It can be used to uncover the interrelationships between the layers of data in a population to inform patient selection or biomarker discovery. We present an application of our framework to non-small cell lung cancer (NSCLC) to identify and separate communities of patients based on their survival and identify the associated biomarkers. We identified potentially mislabeled patients that do not share all the characteristics of the cancer subtype to which they are assigned: either lung adenocarcinoma (LUAD) or lung squamous cell carcinoma (LUSC). Crucially, the community-based biomarkers for poor or long survivors were validated on the whole population. Methods: Our KG framework was leveraged by applying supervised community detection to NSCLC data from TCGA, specifically RNA expression and DNA methylation, with overall survival (OS) as the endpoint (n=999 subjects). Biomarkers associated to each community were ranked based on their prevalence inside against their prevalence outside the community. Results: We obtained 3 communities (391 + 229 + 379 patients) that are all significantly separated by their OS (p<0.05). While KG-derived communities largely overlapped with histology-labelled LUAD or LUSC (concordant LUAD (n=330)/concordant LUSC (n=342)), a small number of patients did not (discordant LUAD (n=37)/discordant LUSC (n=61)). Discordant LUAD patients had significantly lower OS than concordant LUAD (p=0.0198), despite being both labeled as LUAD. Many of the discordant LUSC lacked the 3q26 amplification commonly seen in LUSC and other squamous cell carcinomas. The results from our KG framework highlight its increased sensitivity in relation to existing tools (Cline, Sci Rep 3, 2652 (2013)) as we identified a discordant LUAD group in addition to a discordant LUSC group. Moreover, our tool can select the biomarkers most prevalent in each community, and these significantly separated long from poor survivors on the whole population (p=3.53e-5). Some of these KG-identified biomarkers are known regulators of progression and survival in NSCLC, whereas others are still not extensively studied. This highlights another advantage of our approach in prospective target discovery. Conclusions: Our KG framework allowed the observation of potential misclassifications of tumor subtypes in NSCLC (TCGA). This approach is a proof-of-concept of the value KGs have in identifying signals in muti-omic data that may improve patient stratification and uncover associated biomarker signatures. Our flexible end-to-end framework can take any type of ‘omic data and can be applied to any tumor type for such findings, which represents an alternative to pre-defined KG architectures with defined relationships. Citation Format: Miguel Goncalves, Jake Cohen-Setton, Ioannis Kagiampakis, Ben Sidders, Krishna Bulusu. Multi-modal knowledge graphs enhance patient stratification & biomarker discovery [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4888.

Abstract 1463: Characterizing inflammation-induced epigenetic and transcriptomic alterations driving NSCLC using lung organoid models

Abstract Chronic inflammation is a major driver of lung cancer and contributes to its pathogenesis by inducing genetic and epigenetic alterations. These, in turn, modulate the microenvironment to facilitate emergence of tumor promoting immune evasive mechanisms. The evolution of epigenetic events and their role, both alone, and in association with lung cancer specific genetic drivers in initiation of lung cancer is not well delineated. Specific pathological subtypes of non-small cell lung cancer (NSCLC) have distinct cells of origin, suggestive of cell type-specific changes in driving their development. We employed lung organoid models with chronic exposure to cigarette smoke condensate (CSC) to elucidate sequential, cell type-specific epigenetic and transcriptomic alterations and their role in co-operating with genetic events to drive development of NSCLC. We examined epigenomic and transcriptomic alterations over a 6-month exposure period by DNA methylation arrays, ATAC-seq, and RNA-seq. We observe that CSC exposure causes distinct morphological changes in organoid structure and composition, accompanied by an increase in proliferative potential. Using flow cytometry and immunofluorescent staining approaches, we demonstrate a chronic inflammation-driven shift in stem cell populations accompanied by reduction in differentiated cell types. Bulk and single-cell RNA sequencing approaches identified distinct temporal changes in transcriptional signatures in CSC exposed organoids suggestive of evolution of an immune evasive and pro-tumorigenic phenotype. Gene set enrichment analysis revealed a downregulation of interferon signaling and upregulation of Myc targets and oxidative phosphorylation pathways in CSC treated organoids. Notably, co-culturing organoids with key immune cell populations, demonstrate the ability of the CSC exposed organoids to modulate shifts in immune cells from a pro- to anti- inflammatory state. In accordance with this shift in immune cell phenotype, CSC treated organoids secreted decreased levels of pro-inflammatory cytokines as measured by multiplex ELISA assays. Genome wide DNA methylation analysis reveal that CSC methylated genes were associated with differentiation and cell fate commitment as well as transcription factor activity. Finally, we demonstrate that introduction of key NSCLC-specific driver events at the 6-month time point sensitize only CSC treated organoids to tumor formation in vivo, leading to the development of distinct genetic driver-specific subtypes of NSCLC. Altogether, our results demonstrate the ability of chronic inflammation to drive key cell type-specific transcriptomic and epigenetic changes leading to the development of distinct subtypes of NSCLC. Our findings will help identify molecular correlates for early detection and aid development of novel therapeutic strategies for disease interception. Citation Format: Na Wang, Kara Lombardo, Raksha Padaki, Ray-Whay Chiu Yen, Malcolm V. Brock, Leslie Cope, Hariharan Easwaran, Stephen B. Baylin, Michelle Vaz. Characterizing inflammation-induced epigenetic and transcriptomic alterations driving NSCLC using lung organoid models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1463.

Abstract 986: Mathematical modeling of ctDNA dynamics to predict utility of ultrasensitive MRD assays in early stage non-small cell lung cancer (NSCLC)

Abstract Background: ctDNA assay use for minimal residual disease (MRD) detection in early stage NSCLC has been limited by suboptimal sensitivity of first generation assays. It is unclear how improving assay limit of detection (LOD) would translate to improvements in clinical sensitivity. Here, we performed mathematical modeling of ctDNA dynamics after curative-intent treatment in early stage NSCLC to predict the clinical impact of MRD assays with improved analytical sensitivity. Methods: We analyzed a recently reported dataset of post-surgical ctDNA MRD surveillance in early stage NSCLC (Abbosh et al., Nature 2023) to investigate ctDNA growth dynamics. Of 70 relapsed patients in this cohort, 23 had 3 consecutive ctDNA-positive samples with no intervening therapy. We used these to generate patient-specific mathematical models for ctDNA variant allele fractions (VAF) over time, initially assuming exponential growth. We assessed this assumption by creating log-linear models for ctDNA increase in these cases. Using slopes of the log-linear models, we calculated patient-specific doubling times and assessed their relation to stage, histology, and outcome. We used the log-linear models to simulate a VAF distribution each day after surgery, enabling projection of clinical sensitivities and estimation of lead times for recurrence detection at varying LODs. Results: For the 23 cases with 3+ consecutive ctDNA positive samples, we developed log-linear models for ctDNA VAF growth, which were strongly correlated (r > 0.5) in 18/23 (78%) patients (median r = 0.89), supporting that ctDNA can be modeled as exponential growth. We calculated ctDNA doubling times for the 18 patients with exponential ctDNA kinetics and 10 patients with 2 consecutive ctDNA positive samples. Doubling times were remarkably consistent with median 51 days (IQR 40-79). Doubling times were longer for stage I tumors compared to stage II/III (p = 0.04) but were not different between NSCLC subtypes. Doubling time correlated with time to clinical relapse, with patients having more rapid ctDNA growth relapsing earlier (r = 0.52, p=0.03). Finally, we used patient-specific log-linear models to project ctDNA VAFs at a landmark 30 days after surgery. By improving the LOD for ctDNA detection from 10−4 to 10−6, the projected clinical sensitivity doubled (from 43% to 86%). Simulations predicted that this improvement in sensitivity would increase median lead time from 183 days for LOD=10−4 to 349 days for LOD=10−6 (p<0.0001). Conclusion: For most relapsed NSCLC patients, the natural history of ctDNA MRD growth can be reliably modeled with exponential growth kinetics. ctDNA doubling time is associated with stage and clinical outcome. Ultrasensitive MRD assays with LOD of 10−6 are predicted to meaningfully improve clinical sensitivity at the post-surgical landmark and to increase the lead time for MRD detection before relapse. Citation Format: Jordan S. Goldstein, Christopher Abbosh, Ash A. Alizadeh, Maximilian Diehn, David M. Kurtz. Mathematical modeling of ctDNA dynamics to predict utility of ultrasensitive MRD assays in early stage non-small cell lung cancer (NSCLC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 986.

Abstract 2607: Dual payload immunochemotherapy to treat CD276 positive cancers

Abstract The Cancer Genome Atlas (TCGA) dataset showed higher level of CD276 (B7-H3) transcript in multiple cancers, including triple-negative breast cancers (TNBCs), non-small cell lung cancers (NSCLCs) and other cancers, than normal tissues. Our immunohistochemistry staining of hundreds of patient tissue microarrays confirmed the high expression of CD276 surface receptor in >60% of TNBC and >70% of NSCLC while minimal or low expression in 33 normal organs or tissues. This study aims to develop and evaluate a novel treatment strategy, i.e. anti-CD276 mAb-dual payloads conjugate (DualADC), as immune-chemotherapy to treat CD276-positive TNBCs and NSCLCs. First, a new CD276-targeting monoclonal antibody (mAb) capable of delivering payloads and upregulating tumoral immunity was developed and engineered. Then, a new platform to concurrently conjugate both traditional cytotoxic payload and immunoregulating regent with one CD276 mAb (DualADC) was established. Our evaluations showed that this targeted therapy effectively killed multiple subtypes of TNBC and NSCLC, significantly upregulated immune functions in tumor microenvironment, and reduced tumor burden by up to 90-100% in seven animal models (including patient-derived xenograft). The post-treatment analysis using single-cell RNA sequencing, Luminex multiplex cytokine assay, whole blood, histology and histopathology analysis demonstrated an integrated anti-cancer mechanisms and minimal systematic toxicity. Altogether, our study suggested that the 276-DualADC could provide a promising targeted chemo-immunotherapy for CD276-postive cancer patients. In future, translational investigations will be performed to determine the toxicology, optimal dose, and treatment strategy, and the combination with other therapy will also be evaluated. Citation Format: Zhuoxin Zora Zhou, Jiashuai Zhang, Lufang Zhou, Xiaoguang Margaret Liu. Dual payload immunochemotherapy to treat CD276 positive cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2607.

Abstract 1536: Deciphering the crosstalk within the tumor microenvironment of NSCLC by a virtual microdissection approach

Abstract Introduction: Non-small cell lung cancer (NSCLC) is a highly heterogenous disease with the largest number of cancer-related mortality worldwide, one of the reasons for this is the complex and diverse tumor microenvironment (TME) comprising of numerous cell types. Several studies have already highlighted the importance of TME in dictating progression steps and response to therapies; however, a transcriptome-based molecular subtyping of patients in lung adenocarcinomas (LUADs) and lung squamous cell carcinomas (LUSCs) can further determine the distinct tumor immune microenvironment (TiME), which can eventually provide a systematic overview to improve the diagnosis and prognosis of patients. Material and method: To elucidate such nature of interactions between tumor cells and cells comprising the TME, we exploited the transcriptome of 300 early stages (Ib-IIIa) NSCLC recruited in the prospective observational clinical trial PROMOLE. With the help of a clustering approach, initially we performed a molecular-based virtual stratification/dissection on the NSCLC patients. Next, to elucidate the relative cell-type abundance, a deconvolution approach was applied to identify the possibility of tumor infiltrating immune cells within these subgroups. Immunohistochemistry (IHC) was then used to substantiate these predictions on tumor cells. Results and discussion: The resulting subgroups of LUADs and LUSCs are biologically well-characterized by mutational and gene expression profiles. Cell-type abundance approach identified samples which are enriched with tumor infiltrating immune cells like Neutrophils, Tcells, macrophages, etc. These findings were positively confirmed by IHC with multiple cell markers such as MPO, CD4, CD8, CD68, etc. Integrating these two results highlighted the proportion of TiME in the two different sub-populations along with shedding some light on the crosstalk happening between different cancer-/immune- cell lines. Conclusion: The in-silico predictions on bulk RNA data by virtual micro-dissection, distinguished the two distinct NSCLC subtypes, each associated with clinical and molecular features. Furthermore, the immune cells infiltration suggests a possible role of infiltrating tumor immune cells with the prognosis of patients. Our analysis successfully performed an intra-sample and inter-sample comparison, which can unveil new prognostic markers that can provide relevant information for cancer immunotherapy. Citation Format: Sushant Parab, Francesca Napoli, Davide Corà, Gabriella Doronzo, Valentina Communanza, Luisella Righi, Luca Primo, Valentina Monica, Lorenzo Manganaro, Bianco Selene, Paolo Bironzo, Giorgio Scagliotti, Federico Bussolino. Deciphering the crosstalk within the tumor microenvironment of NSCLC by a virtual microdissection approach [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1536.

Abstract 2750: Leveraging machine-learning approaches to dissect drivers of clinical metastatic dynamics in lung adenocarcinoma

Abstract Background: Non-small cell lung cancer (NSCLC) survival is closely linked to metastatic progression. 5-year survival rates drop from >65% when localized to ~9% when distant metastases occur. Despite this impact on mortality, drivers of overall metastasis and site-specific organotropism are not well understood in NSCLC. To investigate these questions, we focused on Lung adenocarcinoma (LUAD), the most common subtype of NSCLC. Although previous efforts have been made to examine LUAD organotropism, we sought to expand upon longitudinal modeling approaches and explore associations in an independent large clinical and genomic annotated cohort. Methods: Using the artificial intelligence model from (Kehl et al. 2021), we annotated a clinico-genomic cohort of 2777 patients with lung adenocarcinoma, resulting in 59,177 annotated imaging reports. This resulted in time to metastatic site annotations for each patient, with median of 2 years follow up. Metastatic sites included brain, bone, adrenal, liver, lymph node, and mesentery. All patients were evaluated at the Dana-Farber Cancer Institute, and each patient had at least one tumor biopsy sequenced with targeted panel sequencing of 400+ cancer related genes. To infer genomic and clinical correlates of site-specific metastasis, we performed cause-specific and Fine-Gray hazard modeling in this cohort. We used the recurrent event Andersen-Gill model to infer predictors of longitudinal metastasis rates. Results: Site-specific analyses yielded genes whose mutant status were significantly associated with a change in risk and/or rate of certain metastatic sites. We found that NOTCH1 is significantly associated with an increased rate and risk of brain metastases (HR=2.46, P<0.001). SETD2 is associated with decreased incidence of brain (HR=0.42, P<0.001). SMARCA4 is significantly associated with increased incidence and rate of adrenal metastasis (HR=1.64, P<0.01), while NF1 is significantly associated with decreased rate of adrenal metastasis (HR=0.72, P<0.01). In a multivariable model, Female patients and younger patients had a lower rate of new metastatic sites. As previously seen, TP53 and SMARCA4 mutations associated with higher rates of new metastatic sites (HR=1.17, P<0.001; HR=1.36, P<0.001 respectively), and SETD2 is associated with a lower rate of new metastatic sites (HR=0.71, P<0.001). Conclusions: By using time-to-event analysis with longitudinal metastatic annotations, we identify potential drivers of overall metastasis and site-specific organotropism in LUAD patients. Identification of new associations and concordance of our results with previous studies also supports the utility of artificial-intelligence-annotated datasets, allowing for larger cohorts without the need for manual annotation. Citation Format: Tyler Aprati, Michael Manos, Giuseppe Tarantino, Marc Glettig, Maryclare Griffin, Alexander Gusev, Kenneth Kehl, David Liu. Leveraging machine-learning approaches to dissect drivers of clinical metastatic dynamics in lung adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2750.

Abstract 3499: Non-small cell lung cancer (NSCLC) histology classification using DNA methylation data captured from liquid biopsies

Abstract Background: Non-small cell lung cancer (NSCLC) makes up ~85% of lung cancers, and its major subtypes are adenocarcinoma (AD) and squamous cell carcinoma (SC). The histology classification of NSCLC has significant therapeutic and prognostic implications. As liquid biopsies are rapid and non-invasive, computational methods to classify NSCLC subtypes using cell-free DNA (cfDNA) are desired. Methods: Using The Cancer Genome Atlas (TCGA) NSCLC methylation data, we developed a method to classify AD and SC subtypes from cfDNA samples. We mapped TCGA methylation data to Guardant Infinity’s promoter regions and transformed them to allow for the transfer of the TCGA trained model to Infinity cfDNA epigenomic data. An advantage of this approach is that TCGA training data originates from the primary tumor, while cfDNA samples have low tumor fraction (TF) and may also contain signals from metastatic tumors. We trained a penalized Logistic Regression model with LASSO. To better fit the model to Infinity cfDNA data, we performed region filtering. Results: Our model effectively classified AD and SC subtypes from TCGA samples (AUC 0.97) (10-fold cross validation; Table 1). When evaluating on cfDNA samples with epigenomic TF ≥ 1%, the final model shows reasonable performance (AUC 0.94) (Table 1); the performance decreases when reduced to TF > 0.05% (AUC 0.89) as expected (Table 1). When evaluating cfDNA from uncategorized lung samples, the predicted AD probabilities have significant association with KRAS mutations (p-value = 0.027). Conclusion: Our results show that the penalized Logistic Regression model trained with tissue TCGA methylation data through mapping, transformation, and region selection can classify NSCLC subtypes from cfDNA samples. This method may offer a fast, non-invasive tool to inform subtype-based clinical decisions, especially in the setting of poorly-differentiated carcinomas, and to identify subtype switches upon resistance to targeted therapies. Table 1: Model evaluation on tissue TCGA samples and cfDNA Infinity samples (different TF cutoffs) Sample type N samples AUC Accuracy Sensitivity Specificity Precision F1 score tissue TCGA 907 0.97 0.93 0.94 +/- 0.02 0.93 +/- 0.02 0.94 +/- 0.02 0.94 +/- 0.01 cfDNA Infinity 17 (TF ≥ 1%) 0.94 0.88 0.83 +/- 0.3 0.91 +/- 0.17 0.83 +/- 0.3 0.83 +/- 0.21 cfDNA Infinity 29 (TF > 0.05%) 0.89 0.83 0.88 +/- 0.23 0.81 +/- 0.17 0.64 +/- 0.28 0.74 +/- 0.2 Citation Format: Laura Tung, Denis Tolkunov, Catalin Barbacioru, Deepika Kale, Scott Shell, Jill Tsai, Leylah Drusbosky, Kaushal Parikh, Han-Yu Chuang. Non-small cell lung cancer (NSCLC) histology classification using DNA methylation data captured from liquid biopsies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3499.

Abstract 2151: Additional impact of genetic ancestry over self-reported race/ethnicity to prevalence of KRAS mutations and allele-specific subtypes in non-small cell lung cancer

Abstract Background: KRAS mutation is the most common oncogenic driver in patients with non-small cell lung cancer (NSCLC). However, the interplay between self-reported race and/or ethnicity (SIRE) and genetically inferred ancestry (GIA) on KRAS mutations and allele-specific subtypes is largely unknown. Methods: A total of 3918 NSCLC patients from the Chinese OrigiMed (OM) cohort (n=2039) and Boston Lung Cancer Survival (BLCS) cohort (n=1879) were included, all of whom had baseline covariates including age, sex, clinical stage, SIRE, histology and smoking status. KRAS mutation and subtypes were determined by targeted NGS panels in each cohort, and KRAS mutation outcomes, including KRAS mutation, KRAS subtypes, and KRAS transversion and transition were analyzed. GIA was inferred from paired germline data for patients in the BLCS cohort. Multivariable and multinomial logistic regressions were employed to assess the associations between SIRE, GIA and KRAS outcomes. Results: In the entire cohort, 804 (20.5%) patients harbored KRAS mutation, with a considerably higher prevalence of 31.5% observed in BLCS compared to 10.4% in OM cohort. Patients with self-identified Asian was associated with a significantly lower rate of KRAS mutation (OR = 0.44, 95% CI:0.22 - 0.81, P = 0.01), transversion substitutions (OR = 0.24, 95% CI:0.10 - 0.62, P = 0.003) and KRASG12C (OR = 0.17, 95% CI:0.04 - 0.72, P = 0.02) compared to SIRE-White patients after adjusting for confounders including smoking status, sex and histology. The association of SIRE and KRAS driver mutation was further modified by sex, with Asian women demonstrating lower risk compared to Asian men, while the opposite trend was observed in SIRE-White patients. Increased European ancestry was associated with a higher risk of KRAS mutation (OR per 10% increase = 1.05, 95% CI:1.00 - 1.09, P = 0.03), particularly with more transition substitutions (OR per 10% increase = 1.10, 95% CI:1.01 - 1.21, P = 0.03) and KRASG12D (OR per 10% increase = 1.18, 95% CI:1.04 - 1.34, P = 0.01). This association was even more pronounced in patients of increased Northern European ancestry. Furthermore, even among SIRE-White patients, a 10% increase in European ancestry was associated with an additional 6% increase in the likelihood of KRAS mutations (OR = 1.06, 95% CI: 1.01 - 1.12, P = 0.03), while a 10% increase in Admixed American ancestry was associated with a 6% decreased likelihood of KRAS mutations (OR = 0.94, 95% CI: 0.88 - 0.99, P = 0.04). Results were validated in an independent cohort of 1450 patients. Conclusion: Self-reported race/ethnicity and genetic ancestry are both associated with KRAS driver mutations and allele-specific subtypes of NSCLC. This work highlights the significance of integrating both into population-based cancer research, aiming to refine clinical decision-making processes and mitigate health disparities. Citation Format: Xinan Wang, Kangcheng Hou, Biagio Ricciuti, Joao V. Alessi, Xihao Li, Federica Pecci, Rounak Dey, Jia Luo, Mark M. Awad, Alexander Gusev, Xihong Lin, Bruce E. Johnson, David C. Christiani. Additional impact of genetic ancestry over self-reported race/ethnicity to prevalence of KRAS mutations and allele-specific subtypes in non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2151.

Abstract 1534: The tumor immune microenvironment at different stages of non-small cell lung cancer

Abstract Introduction: To better understand the immune microenvironment in non-small cell lung cancer (NSCLC), we evaluated immune cell populations within the tumor parenchyma and their interaction at the periphery of the tumors at different stages of disease. Methods: 27 whole tumor slides were collected from NSCLC patients of all stages after surgical resection. Tissue slides were stained with multiplex Opal immunofluorescence for CD8, CD68, CK, PD-1, PD-L1, FoxP3, and DAPI using a Leica Bond automated stainer. Slides were scanned at 20x magnification by Akoya Polaris whole slide scanner and spectrally unmixed. Spatial analyses were performed on the HALO software. Results: Density of CD8 T cells, T regulatory (Treg) cells, and macrophages were significantly higher (p=0.001) in the interface region compared to the tumor and stroma. There is abundant CD8 T cell infiltration in the tumor at early stage I and this decreased significantly (p=0.05) at higher stages. PD-1− CD8 T cells significantly dropped with higher stage, while PD-1+ CD8 T cells did not. On the other hand, macrophage infiltration increased as stage increased, with most of the increase driven by PD-L1− macrophages. No statistically significant trends were observed in Treg cells across stages, but they appeared to be higher in stage II. No differences in immune infiltration were observed between the NSCLC subtypes adenocarcinoma and squamous cell carcinoma. No correlation was found between proximity of immune cell subsets to tumor cells and stage. Conclusion: The tumor microenvironment in NSCLC is rich in CD8+ cells at early stage and their density decreases as the stage of the disease increases, while macrophage infiltration increases as the disease stage progresses. Citation Format: Mackenzie Hansen, Qin Zhou, Yoko S. DeRose, Magdalena Kovacsovics, Benjamin Brintz, Benjamin L. Witt, Kim L. O'Neill, Siwen Hu-Lieskovan. The tumor immune microenvironment at different stages of non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1534.

Prediction uncertainty estimates elucidate the limitation of current NSCLC subtype classification in representing mutational heterogeneity

The heterogeneous pathogenesis and treatment response of non-small cell lung cancer (NSCLC) has led clinical treatment decisions to be guided by NSCLC subtypes, with lung adenocarcinoma and lung squamous cell carcinoma being the most common subtypes. While histology-based subtyping remains challenging, NSCLC subtypes were found to be distinct at the transcriptomic level. However, unlike genomic alterations, gene expression is generally not assessed in clinical routine. Since subtyping of NSCLC has remained elusive using mutational data, we aimed at developing a neural network model that simultaneously learns from adenocarcinoma and squamous cell carcinoma samples of other tissue types and is regularized using a neural network model trained from gene expression data. While substructures of the expression-based manifold were captured in the mutation-based manifold, NSCLC classification accuracy did not significantly improve. However, performance was increased when rejecting inconclusive samples using an ensemble-based approach capturing prediction uncertainty. Importantly, SHAP analysis of misclassified samples identified co-occurring mutations indicative of both NSCLC subtypes, questioning the current NSCLC subtype classification to adequately represent inherent mutational heterogeneity. Since our model captures mutational patterns linked to clinical heterogeneity, we anticipate it to be suited as foundational model of genomic data for clinically relevant prognostic or predictive downstream tasks.

Influence of TP53 Comutation on the Tumor Immune Microenvironment and Clinical Outcomes With Immune Checkpoint Inhibitors in STK11-Mutant Non-Small-Cell Lung Cancer.

Non-small-cell lung cancer (NSCLC) with STK11mut has inferior outcomes to immune checkpoint inhibitors (ICIs). Using multiomics, we evaluated whether a subtype of STK11mut NSCLC with a uniquely inflamed tumor immune microenvironment (TIME) harboring TP53 comutations could have favorable outcomes to ICIs. NSCLC tumors (N = 16,896) were analyzed by next-generation sequencing (DNA-Seq/592 genes). A subset (n = 5,034) underwent gene expression profiling (RNA-Seq/whole transcriptome). Exome-level neoantigen load for STK11mut NSCLC was obtained from published pan-immune analysis. Tumor immune cell content was obtained from transcriptome profiles using the microenvironment cell population (MCP) counter. ICI data from POPLAR/OAK (n = 34) and the study by Rizvi et al (n = 49) were used to model progression-free survival (PFS), and a separate ICI-treated cohort (n = 53) from Dana-Farber Cancer Institute (DFCI) was used to assess time to treatment failure (TTF) and tumor RECIST response for STK11mutTP53mut versus STK11mutTP53wt NSCLC. Overall, 12.6% of NSCLC tumors had a STK11mut with the proportions of tumor mutational burden (TMB)-high (≥10 mut/Mb), PD-L1 ≥50%, and microsatellite instability-high being 38.3%, 11.8%, and 0.72%, respectively. Unsupervised hierarchical clustering of STK11mut (n = 463) for stimulator of interferon-gamma (STING) pathway genes identified a STING-high cluster, which was significantly enriched in TP53mut NSCLC (P < .01). Compared with STK11mutTP53wt, tumors with STK11mutTP53mut had higher CD8+T cells and natural killer cells (P < .01), higher TMB (P < .001) and neoantigen load (P < .001), and increased expression of MYC and HIF-1A (P < .01), along with higher expression (P < .01) of glycolysis/glutamine metabolism genes. Meta-analysis of data from OAK/POPLAR and the study by Rizvi et al showed a trend toward improved PFS in patients with STK11mutTP53mut. In the DFCI cohort, compared with the STK11mut TP53wt cohort, the STK11mutTP53mut tumors had higher objective response rates (42.9% v 16.7%; P = .04) and also had longer TTF (14.5 v 4.5 months, P adj = .054) with ICI. STK11mut NSCLC with TP53 comutation is a distinct subgroup with an immunologically active TIME and metabolic reprogramming. These properties should be exploited to guide patient selection for novel ICI-based combination approaches.

Data mining-based identification of epigenetic signatures with discrimination potential of lung adenocarcinoma and squamous cell carcinoma.

Mounting evidence suggests that lung adenocarcinoma (LAC) and lung squamous cell carcinoma (LSC) have different biological behaviors and therapeutic regimens in clinical practice. However, limited improvements in molecular differential diagnosis of the two entities have been achieved in recent decades. We aimed to find novel markers that could define non-small cell lung cancer (NSCLC) subtypes. We first explored publically available databases to search for DNA methylation signatures that enable a precise discrimination of LAC and LSC. Next-generation sequencing (NGS) was then used to analyze the methylation status and sites of candidate genes in LAC/LSC tissue samples, and a quantitative methylation-sensitive PCR (qMS-PCR) assay was conducted to test the performance of the selected maker in tissue samples and bronchoalveolar lavage fluid (BALF) specimens. We screened 19 top-ranked methylation loci that are differentially methylated between LAC and LSC. Among these hits, 6 methylation sites are enriched within the PREX1 gene promoter, thus becoming our focus. NGS analysis confirmed markedly higher PREX1 methylation levels in LAC than in LSC and revealed the right sites for detection of PREX1 methylation. Furthermore, PREX1 methylation analysis in lung cancer tissue samples defined 9 of 11 pathologically proven LACs, as well as 12 of 14 LSCs. In addition, ~ 80% LAC BALF samples showed methylated PREX1 compared to substantially lower test positivity (0-9%) of it in LSC and other lung conditions (P < 0.01). Our pilot study identified a unique epigenetic signature that could effectively distinguish LAC from LSC in various lung samples. It may enhance our in-depth understanding of the biology of lung cancer and pave the way for better accurate diagnosis and treatment stratification in the future.

Lack of imbalance between the master regulators TTF1/NKX2-1 and ΔNp63/p40 implies adverse prognosis in non-small cell lung cancer

The transcription factors TTF1/NKX2-1 and ΔNp63/p40 are the counterposed molecular markers associated with the main Non-Small Cell Lung Cancer subtypes: TTF1 for adenocarcinoma, p40 for squamous cell carcinoma. Although they generally display a mutually exclusive expression, some exceptions exist simultaneously lacking or (very rarely) expressing both markers, either pattern being associated to poor prognosis. Hence, we quantitatively analyzed the relationship between their coordinated activity and prognosis. By analyzing the respective downstream transcriptional programs of the two genes, we defined a simple quantitative index summarizing the amount of mutual exclusivity between their activities, called Mean Absolute Activity (MAA). Systematic analysis of the MAA index in a dataset of 1018 NSCLC samples replicated on a validation dataset of 275 showed that the loss of imbalance between TTF-1 and p40 corresponds to a steady, progressive reduction in both overall and recurrence-free survival. Coherently, samples correspondent to more balanced activities were enriched for pathways related to increased malignancy and invasiveness. Importantly, multivariate analysis showed that the prognostic significance of the proposed index MAA is independent of other clinical variables including stage, sex, age and smoke exposure. These results hold irrespectively of tumor morphology across NSCLC subtypes, providing a unifying description of different expression patterns.

Plasma Lipidomics Profiling to Identify the Biomarkers of Diagnosis and Radiotherapy Response for Advanced Non-Small-Cell Lung Cancer Patients.

Advanced lung cancer that contributes to a heavy burden on medical institutions is the leading cause of cancer-related death and is often accompanied by metabolic disorders. In this study, we aimed to explore the biomarkers of diagnosis and radiotherapy response in non-small-cell lung cancer (NSCLC) patients by plasma lipidomics analysis. Using triple-quadrupole mass spectrometer analysis, our research characterized the plasma lipid metabolomics profile of 25 healthy controls and 31 advanced NSCLC patients in each of three different radiotherapy phases. The results showed altered lipid elements and concentrations among NSCLC patients with different radiotherapy phases, NSCLC subtypes, and different radiotherapeutic responses. We found that compared to the healthy controls, myelin-associated glycoprotein (MAG), phosphatidylinositol (PI), and phosphatidylserine (PS) were mainly and significantly altered lipid elements (> twofold, and p < 0.05) among NSCLC patients with different radiotherapy phases. Through comparison of lipid elements between bad and good responses of NSCLC patients with radiotherapy, the obviously declined phosphatidylglycerol (PG 18 : 0/14 : 0, 18 : 1/18 : 3, and 18 : 0/20 : 1) or markedly elevated PI (20 : 0/22 : 5 and 18 : 2/22 : 4) and phosphatidic acid (PA 14 : 0/20 : 4, 14 : 0/20 : 3, and 18 : 2/22 : 4) could indicate poor therapeutic response for NSCLC patients. The results of ROC curve analysis suggested that PG (18 : 0/20 : 1 and 18 : 0/14 : 0) could clearly predict the radiotherapeutic response for NSCLC patients, and PS (18 : 0/20 : 0) and cholesterol were the first two lipid components with the most potential for the diagnosis of advanced NSCLC. Our results indicated that plasma lipidomics profiling might have a vital value to uncover the heterogeneity of lipid metabolism in healthy people and advanced NSCLC patients with different radiotherapy phase, and further to screen out radiotherapeutic response-specific biomarkers.

Exploration of oxidized phosphocholine profile in non-small-cell lung cancer.

Introduction: Lung cancer is one of the most frequently studied types of cancer and represents the most common and lethal neoplasm. Our previous research on non-small cell lung cancer (NSCLC) has revealed deep lipid profile reprogramming and redox status disruption in cancer patients. Lung cell membranes are rich in phospholipids that are susceptible to oxidation, leading to the formation of bioactive oxidized phosphatidylcholines (oxPCs). Persistent and elevated levels of oxPCs have been shown to induce chronic inflammation, leading to detrimental effects. However, recent reports suggest that certain oxPCs possess anti-inflammatory, pro-survival, and endothelial barrier-protective properties. Thus, we aimed to measure the levels of oxPCs in NSCLC patients and investigate their potential role in lung cancer. Methods: To explore the oxPCs profiles in lung cancer, we performed in-depth, multi-level metabolomic analyses of nearly 350 plasma and lung tissue samples from 200 patients with NSCLC, including adenocarcinoma (ADC) and squamous cell carcinoma (SCC), the two most prevalent NSCLC subtypes and COPD patients as a control group. First, we performed oxPC profiling of plasma samples. Second, we analyzed tumor and non-cancerous lung tissues collected during the surgical removal of NSCLC tumors. Because of tumor tissue heterogeneity, subsequent analyses covered the surrounding healthy tissue and peripheral and central tumors. To assess whether the observed phenotypic changes in the patients were associated with measured oxPC levels, metabolomics data were augmented with data from medical records. Results: We observed a predominance of long-chain oxPCs in plasma samples and of short-chain oxPCs in tissue samples from patients with NSCLC. The highest concentration of oxPCs was observed in the central tumor region. ADC patients showed higher levels of oxPCs compared to the control group, than patients with SCC. Conclusion: The detrimental effects associated with the accumulation of short-chain oxPCs suggest that these molecules may have greater therapeutic utility than diagnostic value, especially given that elevated oxPC levels are a hallmark of multiple types of cancer.

The assessment of Dickkopf-1 and Dickkopf-2 protein concentration in different subtypes of non-small cell lung cancer subtypes.

Lung cancer is one of the most prevalent cancers worldwide. Dickkopf-1 (DKK-1) and -2 (DKK-2) are important proteins for the regulated Wnt signalling pathway. Alternations in the Wnt pathway are associated with tumour progression. The aim of the study was to analyse the concentration of DKK-1 and DKK-2 in tumour and matched non-tumour (NT) samples of 65 patients with non-small cell lung cancer (NSCLC), including 3 subtypes: adenocarcinoma (AC), squamous cell carcinoma (SCC), and large cell carcinoma (LCC). The protein concentration was measured by enzyme-linked immunosorbent assay (ELISA) in homogenates. The difference between the level of DKK-1 in tumour and NT specimens was not significant for the whole NSCLC group and SCC and LCC subtype, while in AC samples they were significantly higher (p = 0.028). The highest concentration of DKK-1 was found in the advanced NSCLC samples, with the T4 parameter as well as stage III. Significantly decreased DKK-2 concentrations were detected in all NSCLC subtypes (p < 0.05). Moreover, the DKK-2 level was higher in non-smokers than in smokers. The results indicate that concentrations of DKKs were different in relation to subtypes as well as clinical and socio-demographic parameters. The concentration of DKKs could be associated with the progression of NSCLC. We suggest that DKK-1 could play an oncogenic role in AC, while DKK-2 could be a tumour suppressor in all NSCLC subtypes. Dickkopf-1 and DKK-2 proteins could have differential roles in the Wnt signalling pathway, which is important in many cellular processes, such as proliferation and apoptosis.

A Novel Feature Engineering Method Based on Latent Representation Learning for Radiomics: Application in NSCLC Subtype Classification.

Radiomics refers to the high-throughput extraction of quantitative features from medical images, and is widely used to construct machine learning models for the prediction of clinical outcomes, while feature engineering is the most important work in radiomics. However, current feature engineering methods fail to fully and effectively utilize the heterogeneity of features when dealing with different kinds of radiomics features. In this work, latent representation learning is first presented as a novel feature engineering approach to reconstruct a set of latent space features from original shape, intensity and texture features. This proposed method projects features into a subspace called latent space, in which the latent space features are obtained by minimizing a unique hybrid loss function including a clustering-like loss and a reconstruction loss. The former one ensures the separability among each class while the latter one narrows the gap between the original features and latent space features. Experiments were performed on a multi-center non-small cell lung cancer (NSCLC) subtype classification dataset from 8 international open databases. Results showed that compared with four traditional feature engineering methods (baseline, PCA, Lasso and L2,1-norm minimization), latent representation learning could significantly improve the classification performance of various machine learning classifiers on the independent test set (all p<0.001). Further on two additional test sets, latent representation learning also showed a significant improvement in generalization performance. Our research shows that latent representation learning is a more effective feature engineering method, which has the potential to be used as a general technology in a wide range of radiomics researches.

Machine Learning for Prediction of Non-Small Cell Lung Cancer Based on Inflammatory and Nutritional Indicators in Adults: A Cross-Sectional Study.

The aim of this study was to evaluate the potential benefit of blood inflammation in the diagnosis of non-small cell lung cancer (NSCLC) and propose a machine-learning-based method to predict NSCLC in asymptomatic adults. A cross-sectional study was evaluated using medical records of 139 patients with non-small cell lung cancer and physical examination data from May 2022 to May 2023 of 198 healthy controls. The NSCLC cohort comprised 128 cases of adenocarcinoma, 3 cases of squamous cell carcinoma, and 8 cases of other NSCLC subtypes. The correlation between inflammatory and nutritional markers, such as monocytes, neutrophils, LMR, NLR, PLR, PHR and non-small cell lung cancer was examined. Features were selected using Python's feature selection library and analyzed by five algorithms. The predictive ability of the model for non-small cell lung cancer diagnosis was assessed by precision, accuracy, recall, F1 score, and area under the curve (AUC). The results showed that the top 14 important factors were PDW, age, TP, RBC, HGB, LYM, LYM%, RDW, PLR, LMR, PHR, MONO, MONO%, gender. Additionally, the naive Bayes (NB) algorithm demonstrated the highest overall performance in predicting adult NSCLC among the five machine learning algorithms, achieving an accuracy of 0.87, a macro average F1 score of 0.85, a weighted average F1 score of 0.87, and an AUC of 0.84. In feature ranking, platelet distribution width was the most important feature, and the NB algorithm performed best in predicting adult NSCLC diagnosis.

Expression of CXCL8 and its relationship with prognosis in patients with non-small cell lung cancer.

To determine the expression of chemokine 8 (CXCL8) in non-small cell lung cancer (NSCLC) patients and analyze its correlation with tumor characteristics and patient prognosis. We conducted a retrospective analysis of 149 NSCLC patients treated between January 2016 and April 2018, measuring serum CXCL8 expression upon admission or prior to treatment. The clinical characteristics, including lymph node metastasis and staging, based on CXCL8 expression levels, were analyzed. Receiver Operating Characteristic (ROC) curves was drawn to assess its predictive value for lymph node metastasis and staging in NSCLC patients. Furthermore, the Kaplan-Meier curve was plotted to assess the impact of CXCL8 on 5-year survival in NSCLC Patients. NSCLC patients exhibited significantly higher serum CXCL8 levels than those with benign tumors (P<0.001), with the high CXCL8 expression group showing a higher incidence of lymph node metastasis or stage III NSCLC (P<0.01). CXCL8 was identified as an independent predictor of lymph node metastasis (AUC=0.730) and higher TNM stage (AUC=0.708), as well as a validated biomarker for predicting five-year survival in NSCLC patients. This study highlights the strong association between CXCL8 expression in NSCLC and patient prognosis, particularly regarding lymph node metastasis and clinical staging, suggesting the need for further research to explore CXCL8's specific role in the tumor microenvironment and its impact on different NSCLC subtypes.

Navigating challenges in EGFR exon 20 insertions: The PAPILLON trial unravels the promise of amivantamab-chemotherapy

The PAPILLON trial showed that adding amivantamab to carboplatin-pemetrexed for advanced non-small cell lung cancer (NSCLC) patients with EGFR exon 20 insertion mutations is more effective than chemotherapy. Although safety concerns may arise, this study highlights the need for alternative therapeutic strategies beyond chemotherapy for this subtype of NSCLC.