DNA mutations are a well-characterized source of neoepitopes in immunotherapy. Here, we examined the contribution of dysregulated RNA processing to neoantigen production. Leveraging multi-omics and checkpoint inhibitor (CPI) response data from >1,000 patients, we identified reduced activity of the nonsense-mediated mRNA decay (NMD) pathway kinase SMG1 as a predictor of improved CPI response. NMD inhibition through SMG1 targeting stabilized transcripts containing premature termination codons, most of which were of non-mutational origin. This reshaped the major histocompatibility complex class I (MHC class I)-bound immunopeptidome and increased neoantigen abundance to levels comparable to high mutation burden tumors. Functionally, NMD inhibition drove antigen-dependent T cell-mediated tumor cell killing in vitro, promoted activation of tissue-resident T cells in patient-derived models ex vivo, and improved CPI efficacy in vivo. Our findings establish NMD inhibition as a strategy to harness a previously inaccessible source of canonical and non-canonical neoantigens, with the potential to increase tumor immunogenicity across cancers.

Cancers rarely respond completely to immunotherapy. While tumors consist of multiple genetically distinct clones, whether this affects the potential for immune escape remains unclear due to an inability to isolate and propagate individual subclones from human cancers. Here, we leverage the multi-region TRACERx lung cancer evolution study to generate a patient-derived organoid - T cell co-culture platform that allows the functional analysis of subclonal immune escape at single clone resolution. We establish organoid lines from 11 separate tumor regions from three patients, followed by isolation of 81 individual clonal sublines. Co-culture with tumor infiltrating lymphocytes (TIL) or natural killer (NK) cells reveals cancer-intrinsic and subclonal immune escape in all 3 patients. Immune evading subclones represent genetically distinct lineages with a unique evolutionary history. This indicates that immune evading and non-evading subclones can be isolated from the same tumor, suggesting that subclonal tumor evolution directly affects immune escape.

Immunotherapy has revolutionized cancer treatment, yet only a minority of individuals respond clinically, necessitating alternative strategies that can benefit these patients. Novel immuno-oncology targets may achieve this through bypassing resistance mechanisms to standard therapies. We introduce Mining Immunotherapy Drug tArgetS (MIDAS), a multimodal graph neural network system for immuno-oncology target discovery. MIDAS leverages gene interactions, multi-omic patient profiles, immune cell biology, antigen processing, disease associations and phenotypic consequences of genetic perturbations. It generalizes to time-sliced data, outcompetes state-of-the-art baselines (including OpenTargets) and ranks approved targets above those in clinical development. Moreover, MIDAS recovers immunotherapy-response-associated genes in unseen patients, thereby capturing immunotherapy response determinants. Interpretability analyses reveal a reliance on autoimmunity, regulatory networks and immuno-oncology pathways. Functionally perturbing oncostatin M-oncostatin M receptor signalling, a proposed MIDAS target, in TRACERx melanoma-patient-derived explants yielded reduced dysfunctional CD8(+) T cells, which associate with immunotherapy response, and reduced CCL4levels. Furthermore, oncostatin M and oncostatin M receptor expression is associated with altered T cell and macrophage profiles in bulk transcriptomic data from patient samples. These data are consistent with a role for oncostatin M-oncostatin M in modulating the tumour microenvironment towards immunosuppressive, tumour-promoting phenotypes. Our results present a machine learning framework for analysing multimodal data for immuno-oncology target discovery.

Abstract The lung epithelium harbours progenitor populations, including basal, club, neuroendocrine, and alveolar type I/II (AT1/AT2) cells. A highly plastic Krt8+ alveolar intermediate cell state (KAC) arises in Kras-driven tumorigenesis and in lung injury models. Environmental particulate matter (PM) exposure promotes lung tumorigenesis through macrophage-derived interleukin-1β (IL-1β). How EGFR mutations (EGFRm), the most common genomic driver of lung cancer in never-smokers, and tumour promotion drive lung adenocarcinoma (LUAD) at the single cell level from diverse lung epithelial lineages remains unclear. Understanding this will aid development of molecularly targeted prevention approaches, particularly in never-smokers. We generated mouse models of lineage-specific activation of EGFR-L858R within basal, club, neuroendocrine, and AT2 lineages. LUAD development and lineage convergence were assessed using histology and snRNA-seq across 37,627 cells from 100 mice. For PM effects, EGFR-L858R induction in AT2 cells was achieved via Ad5-Spc-Cre intratracheal instillation, followed by PM or PBS exposure. 10X multiome snRNA-seq (34,459 AT2-lineage cells) and snATAC-seq were used to profile transcriptional and epigenetic changes. EGFRm precision-cut lung slices (PCLS) were exposed to PM ex vivo in presence of IgG or anti- IL-1β blocking antibodies. Our analysis revealed that LUAD arises from basal, club, neuroendocrine, and AT2 lineages — all converging on an alveolar-like state, mirroring the trajectory observed in KRAS-driven LUAD. Notably, alveolar KACs were detected in EGFRm cells derived from basal, club, and AT2 lineages, indicating a conserved transitional state during early tumorigenesis. EGFRm activation also induced KAC-signature genes while preserving lineage-specific markers, reflecting a dual programme of lineage convergence and memory. Upon PM exposure, EGFRm KACs exhibited pronounced upregulation of stress-responsive and inflammation-induced epithelial genes. Compared with wild-type AT2 controls, PM-exposed EGFRm KACs exhibited strong enrichment of LUAD-associated transcripts, with over 30% overlap encompassing 1,298 genes. This highlights the synergy between EGFRm and environment-induced injury. Furthermore, we identified epigenetic rewiring of KACs, marked by activation of transcription factors linked to the IL-1β pathway upon PM. Functionally, blocking IL-1β in PCLS effectively inhibited KAC formation, establishing a mechanistic link between inflammatory signalling and early tumour cell fate. In conclusion, we showed that epithelial lineages converge on KACs en-route to EGFR-driven LUAD, which expands and undergoes transcriptional and epigenetic remodelling upon PM-exposure. Future studies defining key regulators of KAC expansion and its IL-1β dependency could inform novel therapeutic avenues for molecular cancer prevention. Citation Format: Michelle M. Leung, Maria Zagorulya, Tej Pandya, Marcellus Augustine, Anthony J. Griffen, Alix Le Marois, Sophia Ward, Hubert Slawinski, Alejandro Suárez-Bonnet, Simon L. Priestnall, Alexandros Hardas, Erik Sahai, Lindsay M. LaFave, Eva Gronroos, Nicholas McGranahan, William Hill, Clare Weeden, Charles Swanton. Air pollutants remodel mutant epithelia toward a convergent lung adenocarcinoma progenitor state [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 931.

Predicting lung cancer risk would enhance prevention trials. Although the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS) trial demonstrated reduced lung cancer incidence with interleukin (IL)-1β inhibition, the high number needed to treat (NNT) to prevent lung cancer limits its use in unselected populations. Using machine learning, we identified a 14-protein plasma signature predicting lung cancer more than 5 years before diagnosis. The signature, validated across eight cohorts, was elevated in current smokers and individuals exposed to particulate matter (PM) and linked to lung myeloid and alveolar cells. In epidermal growth factor receptor (EGFR)-driven lung adenocarcinoma, diverse epithelial lineages converged on a keratin8+/claudin4+ alveolar transitional state (KAC), whose transcriptional programs correlated with signature emergence. Components of the signature were induced by PM, oncogenic EGFR, or IL-1β, whereas IL-1β inhibition restrained PM-driven KAC expansion and early tumorigenesis. In CANTOS, the signature identified individuals who seemed to benefit more from anti-IL-1β therapy, lowering the NNT threshold and nominating circulating signals of tumor promotion for prevention.

Abstract Introduction: Neoantigens from somatic tumor mutations are essential for effective anti-tumor immune responses. Frameshift insertions and deletions (fs-indels) represent a rare but highly immunogenic mutation subtype, as they create novel open reading frames (neoORFs) that generate peptides that are significantly distinct from self-antigens. Nevertheless, fs-indels often introduce premature termination codons, leading to transcript degradation via the nonsense-mediated mRNA decay (NMD) pathway, leading to loss of immunogenic neoantigen. Approach: For the first time, we pharmacologically inhibited SMG1, a core component of the NMD pathway, across a range of preclinical models, including human and mouse cancer cell lines, patient-derived tumor organoids (PDTOs), patient-derived tumor fragments (PDTFs), and syngeneic mouse xenografts. We analyzed the changes in transcriptome, proteome, and immunopeptidome following SMG1 inhibition (SMG1i) and peptide reactivity in in vitro priming experiments. We then combined tumor-T cell co-cultures and PDTFs to assess the anti-tumor immunogenicity induced by SMG1i. Ex vivo and in vivo immunological responses were assessed by high-dimensional flow cytometry, cytometric bead array, and single-cell RNA- and TCR-sequencing. Results: Using multi-omic and checkpoint inhibitor (CPI) response data from over 1,000 patients, we show that decreased expression of the key NMD mediator, SMG1, correlates with improved CPI response. Inhibiting SMG1 ex vivo and in vivo activates and expands tumor-reactive T cells and sensitizes CPI efficacy. Mechanistically, SMG1 inhibition stabilizes frameshift-derived transcripts, increasing the abundance and surface presentation of immunogenic neoantigens. This results in an increase in neoepitope burden in tumors, similar to that seen in tumors with high tumor mutational burden (TMB), without inducing DNA damage. Co-culturing tumor cells and PDTOs with CD8+ T cells after SMG1i results in strong MHC class I antigen-dependent T cell activation and tumor cell killing. Conclusion: Our findings highlight SMG1 inhibition as a promising strategy to exploit an untapped source of highly immunogenic peptides. It enhances anti-tumor immunogenicity without introducing DNA mutations, regardless of tumor type or TMB status, providing translational evidence for sensitizing ICB responses. Citation Format: Hongchang Fu, Roberto Vendramin, Shanila Fernandez Patel, Yue Zhao, Danwen Qian, Lorena Ligammari, Osnat Bartok, Polina Greenberg, Ronen Levy, Andrea Castro, Krupa Thakkar, Jun Murai, Wei-ting Lu, Christopher C. Sng, Chen Weller, Gordon Beattie, Amandeep Bhamra, Roc Farriol-Duran, Despoina Karagianni, Marcellus Augustine, Krijn Djikstra, Christopher L. Pinder, Benjamin S. Simpson, Gordon Weng-Kit Cheung, TRACERx Consortium, Felipe Galvez Cancino, Petra Vlckova, Silvia Surinova, Manuel Rodriguez-Justo, Mansi Shah, Nicholas McGranahan, Jeremy G. Carlton, Eva Camilla Gronroos, Sergio Quezada, James Luke Reading, Samra Turajlic, Yardena Samuels, Charles Swanton, Kevin Litchfield. Nonsense-mediated mRNA decay inhibition augments in vitro, in vivo, and ex vivo anti-tumor immunity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6742.

Abstract Background: Current lung cancer screening programs rely heavily on age and smoking history, excluding never-smokers and those with minimal smoking exposure. Such criteria have a low positive predictive value (PPV), limiting molecular prevention strategies. Our previous work identified interleukin-1β (IL-1β) as a mediator of lung cancer initiation through environmental particulate matter (PM) exposure, suggesting potential targets for therapeutic cancer prevention. Here, we sought to identify circulating signals predictive of lung cancer prior to clinical diagnosis and determine if they were useful for clinical trial stratification of IL-1β therapy. Methods: Using human plasma proteomic data from the UK Biobank (n=48,099 individuals; 375 lung cancer cases), we developed a machine-learning framework to identify proteins predictive of lung cancer diagnosis. We validated this model in eight independent human cohorts (2,176 cases, 54,324 controls). We further analysed plasma proteomic murine data from EGFR-mutant mice exposed to PM as well as from baseline samples from the CANTOS trial which previously had demonstrated reduction of lung cancer incidence with IL-1β inhibition. Results: Our machine-learning approach identified a plasma signature of 14 proteins, predictive of lung cancer diagnosis up to 6 years before clinical detection, significantly outperforming current lung cancer risk models (p<0.01 by de Long’s test). Validation across eight external human cohorts confirmed consistent associations for all proteins. Mouse experiments demonstrated a sustained increase in circulating signature proteins following PM exposure specifically in EGFR-mutant mice, linking environmental PM exposure directly to the alveolar niche as an early tumour-promoting microenvironment. Retrospective analysis of the CANTOS trial showed the protein signature stratified individuals deriving benefit from IL-1β inhibition, reducing the number needed to treat from 1516 to 55. Discussion: Our findings indicate that a circulating plasma signature derived from alveolar niche remodelling and induced by PM and EGFR-driven oncogenesis can effectively identify individuals at high risk of lung cancer two years before clinical onset. The identified proteins may enable targeted stratification for molecular prevention trials. Future research should focus on extending this approach and developing absolute quantification assays to for clinical translation. Citation Format: Tej Pandya, Maria Zagorulya, Michelle M. Leung, Marcellus Augustine, Lydia Y. Liu, Oleg Blyuss, Jincheng Wu, Marc Pelletier, Vernon Burk, Neil Wright, David Muller, Ka Hung Chan, Ekaterina Pazukhina, Marc Gunter, Elizabeth A. Platz, Karl Smith-Byrne, Nuno Rocha Nene, Eva Camilla Gronroos, Nicholas McGranahan, William Hill, Clare Weeden, Charles Swanton. Plasma proteomics for risk prediction of lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7632.

Chronic obstructive pulmonary disease (COPD) is a lung condition characterised by chronic respiratory symptoms due to abnormalities of the airways that causes persistent, often progressive airflow obstruction and independent risk factor for lung cancer. However, the mechanisms linking COPD to lung cancer remain poorly understood, particularly the role of mutual risk factors such as air pollution and smoking. Emerging evidence challenges that environmental carcinogens can be purely mutagenic, and suggests that chronic inflammation also facilitates cancer progression and can be compounded in genetically susceptible individuals. This study aimed to investigate any gene-environment interactions contributing to lung cancer development in patients with COPD. We used machine learning frameworks to solve this problem given the limitations of conventional methods on this high-dimensional and multi-modal data. Using germline SNPs, measurements of air pollution through PM 2.5, and clinical information from the UK Biobank, a cohort of 488, 377 participants, we created models to predict whether patients with COPD would get lung cancer. We tested multiple frameworks optimised for tabular data such as eXtreme Gradient Boosting (XGBoost) and TabNet, in addition to Elastic Net, to be able to independently capture different facets of the data and account for potential non-linear interactions not modelled in linear approaches. As a secondary aim, we focussed on COPD patients with at least a ten-year post-diagnosis survival, minimizing bias from mortality effects on cancer progression. In (n=24, 094) individuals with COPD, 1, 312 progressed to develop lung cancer by the date of censorship. Patients progressing to lung cancer were significantly older at enrolment, had greater exposure to air pollution, and smoked more heavily (all p < 0.001). Using an 80:20 train-test split, XGBoost and Elastic Net models outperformed TabNet (p< 0.01 by DeLong’s test) in predicting lung cancer risk, with XGBoost achieving the highest ROC-AUC on the held-out test set (0.73, 95% CI: 0.72-0.75). Restricting the inclusion criteria to ten-year COPD survivors notably improved prediction accuracy (ROC-AUC = 0.87, 95% CI: 0.85-0.90) and identified SNPs related to lung function and inflammation (such as RAGE) as important risk factor of lung cancer in patients in COPD patients. This study demonstrates that integrating machine learning with epidemiological data enables high-accuracy modelling of lung cancer risk in patents with COPD, emphasizing a potential role of inflammation mediating PM 2.5 gene-environment interactions. These findings underscore the need for research on the biological mechanisms driving inflammation-associated cancer risk in COPD for patient stratification, as well as the potential for applying machine learning frameworks methods to other cancers to explore tumorigenic inflammation. Tej Pandya, Marcellus Augustine, Kevin Litchfield, Nuno Rocha Nene, Charles Swanton. Modelling COPD to lung cancer progression using machine learning frameworks [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7423.

Tumor-infiltrating B cells play a significant role in tumor development, progression, and prognosis, yet a comprehensive classification system is lacking. To address this gap, we present a pan-cancer single-cell RNA sequencing (scRNA-seq) atlas of tumor-infiltrating B and plasma cells across a large sample cohort. We identify key B cell subset signatures, revealing distinct subpopulations and highlighting the heterogeneity and functional diversity of these cells in the tumor microenvironment. We explore associations between B cell subsets and checkpoint inhibitor therapy responses, finding subset-specific effects on overall response. Additionally, we examine B and T cell crosstalk, identifying unique ligand-receptor pairs for specific B cell subsets, spatially validated. This comprehensive dataset serves as a valuable resource, providing a detailed atlas that enhances the understanding of B cell complexity in tumors and opens new avenues for research and therapeutic strategies.

Abstract Up to 50% of cancer patients are diagnosed at a late stage with tumors that are often unresectable, leading to intensive treatments and a preventable loss of life. Whilst multiple assays have been developed to tackle the issue of disease mortality, these approaches have serious limitations. For example, circulating tumor DNA (ctDNA) preferentially detects hard-to-treat tumors and micrometastatic disease, with limited sensitivity and specificity for early-stage malignancy, underscoring an urgent need for novel early detection strategies. Substantial evidence exists to demonstrate that signals related to innate (e.g. pro-inflammatory cytokines) and adaptive (e.g. antibodies and T cell receptors [TCRs]) immune engagement arise early in the development of cancer. The immune system therefore offers potential as an exquisitely sensitive and highly specific intrinsic early detection system for cancer. Here we propose the development of a pan-cancer novel early detection assay, using peripheral blood to measure three key cancer-specific components of the anti-tumor immune response: i) T cell receptor (TCR) sequences, ii) antibody signatures, iii) cytokine markers. Our preliminary data demonstrates that an age and smoking-matched cohort of lung cancer patients can be distinguished from healthy donors using TCR sequencing data processed through a published machine learning approach (p<0.01). In support of this, a pilot study on a custom peptide microarray comprised of lung cancer associated antigens and recurrent neoantigens demonstrated that lung cancer patient samples show significantly higher antibody intensities than healthy donors. Thirdly, using data from a prospective cohort study, we have demonstrated that inflammatory markers can be leveraged to predict future cancer diagnosis, and identified key protein markers for future work. These data provide proof-of-principle for the use of these analytes as a tool for early detection. We are actively curating a unique cohort of samples for this work from a prospective study, Nodule Immunophenotyping Biomarker for Lung Cancer Early Diagnosis (NIMBLE) (NCT05432739), which recruits patients with indeterminate lung nodules that may represent early cancers (>280 patients to date). This malignancy has high mortality rates and a significant need for early detection strategies. Through this, we aim to understand the early immunobiological response to cancer and leverage this information to design a novel multi-parametric immunopredictor assay for early detection of cancer. Citation Format: Evelyn Fitzsimons, Alexander Coulton, Hongchang Fu, Marcellus Augustine, Richard Lee, James Reading, Kevin Litchfield. Integration of innate and adaptive immune signatures for early detection of 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 6090.

Immunotherapy has revolutionised cancer treatment, yet few patients respond clinically, necessitating alternative strategies that can benefit these patients. Novel immune-oncology targets can achieve this through bypassing resistance mechanisms to standard therapies. To address this, we introduce MIDAS, a multimodal graph neural network system for immune-oncology target discovery that leverages gene interactions, multi-omic patient profiles, immune cell biology, antigen processing, disease associations, and phenotypic consequences of genetic perturbations. MIDAS generalises to time-sliced data, outcompetes existing methods, including OpenTargets, and distinguishes approved from prospective targets. Moreover, MIDAS recovers immunotherapy response-associated genes in unseen trials, thus capturing tumour-immune dynamics within human tumours. Interpretability analyses reveal a reliance on autoimmunity, regulatory networks, and relevant biological pathways. Functionally perturbing the OSM-OSMR axis, a proposed target, in TRACERx melanoma patient-derived explants yielded reduced dysfunctional CD8+ T cells, which associate with immunotherapy response. Our results present a machine learning framework for analysing multimodal data for immune-oncology discovery.

Abstract Overview: Immunotherapy has revolutionized cancer treatment. However, existing immune checkpoint inhibitors (CPI) yield low response rates in most cancers, highlighting the need for new therapeutic options. Traditional immune-oncology (IO) target discovery relies on preclinical models, which struggle to recapitulate human tumor complexity, limiting translation potential. Attention is thus directed at harnessing multimodal patient molecular data with modern machine learning (ML) techniques to identify new IO targets which may have higher clinical viability. Approach: We posed IO target discovery as a binary classification task, training ML systems on known candidate drug targets that have progressed to stage I or higher clinical trials. We constructed a rich knowledge graph database to support model development. Graph nodes (n=11,919) comprised genes with edges linking genes involved in n=99,275 protein-protein interactions (PPIs; PMID: 28936969). Genes were labelled with n=6,387 gene-disease associations alongside bulk exome and transcriptome (RNAseq) features from n=1,317 CPI-treated patients (PMID: 33508232). We considered the role of different immune subsets by incorporating cell type-specific PPIs from single cell RNAseq atlases of n=350 samples (PMID: 31786210). To capture how antigen processing affects immunotherapy response, we also examined the immunopeptidome of n=60 patients (PMID: 30556813). Causal data on immune responses to genetic perturbation stemmed from n=7 publicly available CRISPR tumor-T cell co-cultures and n=15,442 SNP-phenotype links. Lastly, we used interpretability analysis to understand drivers of model predictions and elucidate critical genes that influence multiple IO target pathways. Results: Firstly, we developed an ensemble ML approach which achieved test ROC-AUC>0.75. Next, we used a graph-based ML framework which yielded superior performance (test ROC-AUC>0.90). Orthogonal validation confirmed these models can discriminate known targets by trial phase (p<0.001), predict patient response in new CPI trials (p<0.05), and identify genes that rank highly in unseen genome-wide CRISPR screens. Targets from both methods have entered experimental validation in patient-derived explants and organoid-immune co-cultures. Already, we have identified 2 novel targets predicted to relate to macrophage activity. Early data suggest that perturbing them leads to macrophage repolarization. Further validation experiments are ongoing. Conclusions: We will reveal candidate targets, showing that our method is effective at uncovering new IO targets and can identify critical nodes in biological networks that might be attractive hits. In addition, deciphering data types that drive model predictions provides a broader immunobiological understanding of anti-tumor immune responses. Thus, our results endorse using ML and multimodal data for novel IO target discovery. Citation Format: Marcellus Augustine, Nuno Rocha Nene, Krupa Thakkar, Danwen Qian, Evelyn Fitzsimons, Benjamin S. Simpson, Chris Watkins, Charles Swanton, Kevin Litchfield. Identifying new immunotherapy targets using machine learning and ex vivo validation [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 5887.

While the presence of intratumoural microbes has been well described, their role in tumour evolution is unknown. Here we conducted microbial sequencing of 567 samples from non-small cell lung cancer (NSCLC) patients in the TRACERx 421 cohort, spanning matched primary tumour, metastasis and normal adjacent tissue samples. We demonstrate the presence of cancer-associated bacteria shared across multiple primary tumour regions, enriched for common lung pathogens. Using spatially resolved digital histology data in combination with the bacterial load within the tumour, we show that patients with a high bacterial load and low infiltration of lymphocytes have an overall worse survival. Lastly, signature analysis of mutations in the bacterial genome revealed a predominantly endogenous processes contributing to the mutagenesis of intratumoural bacteria in NSCLC. We present a spatially and longitudinally resolved study of the cancer associated bacteriome in lung cancer.

Multiple synchronous lung cancers (MSLCs) constitute a unique subtype of lung cancer. To explore the genomic and immune heterogeneity across different pathological stages of MSLCs, we analyse 16 MSLCs from 8 patients using single-cell RNA-seq, single-cell TCR sequencing, and bulk whole-exome sequencing. Our investigation indicates clonally independent tumours with convergent evolution driven by shared driver mutations. However, tumours from the same individual exhibit few shared mutations, indicating independent origins. During the transition from pre-invasive to invasive adenocarcinoma, we observe a shift in T cell phenotypes characterized by increased Treg cells and exhausted CD8+ T cells, accompanied by diminished cytotoxicity. Additionally, invasive adenocarcinomas exhibit greater neoantigen abundance and a more diverse TCR repertoire, indicating heightened heterogeneity. In summary, despite having a common genetic background and environmental exposure, our study emphasizes the individuality of MSLCs at different stages, highlighting their unique genomic and immune characteristics. Multiple synchronous lung cancers (MSLCs) are a subtype of lung cancer. Here the authors characterise MSLCs using single cell RNA sequencing, single cell TCR sequencing and bulk whole-exome sequencing to investigate the mutations that arise in and are associated with invasive adenocarcinoma development, and immune microenvironment changes in this process.

A complete understanding of how exposure to environmental substances promotes cancer formation is lacking. More than 70 years ago, tumorigenesis was proposed to occur in a two-step process: an initiating step that induces mutations in healthy cells, followed by a promoter step that triggers cancer development 1 . Here we propose that environmental particulate matter measuring ≤2.5 μm (PM 2.5 ), known to be associated with lung cancer risk, promotes lung cancer by acting on cells that harbour pre-existing oncogenic mutations in healthy lung tissue. Focusing on EGFR-driven lung cancer, which is more common in never-smokers or light smokers, we found a significant association between PM 2.5 levels and the incidence of lung cancer for 32,957 EGFR-driven lung cancer cases in four within-country cohorts. Functional mouse models revealed that air pollutants cause an influx of macrophages into the lung and release of interleukin-1β. This process results in a progenitor-like cell state within EGFR mutant lung alveolar type II epithelial cells that fuels tumorigenesis. Ultradeep mutational profiling of histologically normal lung tissue from 295 individuals across 3 clinical cohorts revealed oncogenic EGFR and KRAS driver mutations in 18% and 53% of healthy tissue samples, respectively. These findings collectively support a tumour-promoting role for PM 2.5 air pollutants and provide impetus for public health policy initiatives to address air pollution to reduce disease burden.

Abstract Environmental carcinogenic exposures are major contributors to global disease burden yet how they promote cancer is unclear. Over 70 years ago, the concept of tumour promoting agents driving latent clones to expand was first proposed. In support of this model, recent evidence suggests that human tissue contains a patchwork of mutant clones, some of which harbour oncogenic mutations, and many environmental carcinogens lack a clear mutational signature. We hypothesised that the environmental carcinogen, <2.5μm particulate matter (PM2.5), might promote lung cancer promotion through non-mutagenic mechanisms by acting on pre-existing mutant clones within normal tissues in patients with lung cancer who have never smoked, a disease with a high frequency of EGFR activating mutations. We analysed PM2.5 levels and cancer incidence reported by UK Biobank, Public Health England, Taiwan Chang Gung Memorial Hospital (CGMH) and Korean Samsung Medical Centre (SMC) from a total of 463,679 individuals between 2006-2018. We report associations between PM2.5 levels and the incidence of several cancers, including EGFR mutant lung cancer. We find that pollution on a background of EGFR mutant lung epithelium promotes a progenitor-like cell state and demonstrate that PM accelerates lung cancer progression in EGFR and Kras mutant mouse lung cancer models. Through parallel exposure studies in mouse and human participants, we find evidence that inflammatory mediators, such as interleukin-1ꞵ, may act upon EGFR mutant clones to drive expansion of progenitor cells. Ultradeep mutational profiling of histologically normal lung tissue from 247 individuals across 3 clinical cohorts revealed oncogenic EGFR and KRAS driver mutations in 18% and 33% of normal tissue samples, respectively. These results support a tumour-promoting role for PM acting on latent mutant clones in normal lung tissue and add to evidence providing an urgent mandate to address air pollution in urban areas.

The genetic evolutionary features of solid tumour growth are becoming increasingly well described, but the spatial and physical nature of subclonal growth remains unclear. Here, we utilize 102 macroscopic whole-tumour images from clear cell renal cell carcinoma patients, with matched genetic and phenotypic data from 756 biopsies. Utilizing a digital image processing pipeline, a renal pathologist marked the boundaries between tumour and normal tissue and extracted positions of boundary line and biopsy regions to X and Y coordinates. We then integrated coordinates with genomic data to map exact spatial subclone locations, revealing how genetically distinct subclones grow and evolve spatially. We observed a phenotype of advanced and more aggressive subclonal growth in the tumour centre, characterized by an elevated burden of somatic copy number alterations and higher necrosis, proliferation rate and Fuhrman grade. Moreover, we found that metastasizing subclones preferentially originate from the tumour centre. Collectively, these observations suggest a model of accelerated evolution in the tumour interior, with harsh hypoxic environmental conditions leading to a greater opportunity for driver somatic copy number alterations to arise and expand due to selective advantage. Tumour subclone growth is predominantly spatially contiguous in nature. We found only two cases of subclone dispersal, one of which was associated with metastasis. The largest subclones spatially were dominated by driver somatic copy number alterations, suggesting that a large selective advantage can be conferred to subclones upon acquisition of these alterations. In conclusion, spatial dynamics is strongly associated with genomic alterations and plays an important role in tumour evolution.

Genetic intra-tumour heterogeneity fuels clonal evolution, but our understanding of clinically relevant clonal dynamics remain limited. We investigated spatial and temporal features of clonal diversification in clear cell renal cell carcinoma through a combination of modelling and real tumour analysis. We observe that the mode of tumour growth, surface or volume, impacts the extent of subclonal diversification, enabling interpretation of clonal diversity in patient tumours. Specific patterns of proliferation and necrosis explain clonal expansion and emergence of parallel evolution and microdiversity in tumours. In silico time-course studies reveal the appearance of budding structures before detectable subclonal diversification. Intriguingly, we observe radiological evidence of budding structures in early-stage clear cell renal cell carcinoma, indicating that future clonal evolution may be predictable from imaging. Our findings offer a window into the temporal and spatial features of clinically relevant clonal evolution.