Search

Abstract Assessing fertilized human embryos is crucial for in vitro fertilization, a task being revolutionized by artificial intelligence. Existing models used for embryo quality assessment and ploidy detection could be significantly improved by effectively utilizing time-lapse imaging to identify critical developmental time points for maximizing prediction accuracy. Addressing this, we develop and compare various embryo ploidy status prediction models across distinct embryo development stages. We present BELA, a state-of-the-art ploidy prediction model that surpasses previous image- and video-based models without necessitating input from embryologists. BELA uses multitask learning to predict quality scores that are thereafter used to predict ploidy status. By achieving an area under the receiver operating characteristic curve of 0.76 for discriminating between euploidy and aneuploidy embryos on the Weill Cornell dataset, BELA matches the performance of models trained on embryologists’ manual scores. While not a replacement for preimplantation genetic testing for aneuploidy, BELA exemplifies how such models can streamline the embryo evaluation process.

Systemic lupus erythematosus (SLE) is an autoimmune disease, the pathophysiology and genetic basis of which are incompletely understood. Using a forward genetic screen in multiplex families with SLE, we identified an association between SLE and compound heterozygous deleterious variants in the non-receptor tyrosine kinases (NRTKs) ACK1 and BRK. Experimental blockade of ACK1 or BRK increased circulating autoantibodies in vivo in mice and exacerbated glomerular IgG deposits in an SLE mouse model. Mechanistically, NRTKs regulate activation, migration, and proliferation of immune cells. We found that the patients’ ACK1 and BRK variants impair efferocytosis, the MERTK-mediated anti-inflammatory response to apoptotic cells, in human induced pluripotent stem cell (hiPSC)-derived macrophages, which may contribute to SLE pathogenesis. Overall, our data suggest that ACK1 and BRK deficiencies are associated with human SLE and impair efferocytosis in macrophages.

IntroductionThe 2024 Voice AI Symposium, presented by the Bridge2AI-Voice Consortium, featured deep-dive educational workshops conducted by experts from diverse fields to explore the latest advancements in voice biomarkers and artificial intelligence (AI) applications in healthcare. Through five workshops, attendees learned about topics including international standardization of vocal biomarker data, real-world deployment of AI solutions, assistive technologies for voice disorders, best practices for voice data collection, and deep learning applications in voice analysis. These workshops aimed to foster collaboration between academia, industry, and healthcare to advance the development and implementation of voice-based AI tools.MethodsEach workshop featured a combination of lectures, case studies, and interactive discussions. Transcripts of audio recordings were generated using Whisper (Version 7.13.1) and summarized by ChatGPT (Version 4.0), then reviewed by the authors. The workshops covered various methodologies, from signal processing and machine learning operations (MLOps) to ethical concerns surrounding AI-powered voice data collection. Practical demonstrations of AI-driven tools for voice disorder management and technical discussions on implementing voice AI models in clinical and non-clinical settings provided attendees with hands-on experience.ResultsKey outcomes included the discussion of international standards to unify stakeholders in vocal biomarker research, practical challenges in deploying AI solutions outside the laboratory, review of Bridge2AI-Voice data collection processes, and the potential of AI to empower individuals with voice disorders. Additionally, presenters shared innovations in ethical AI practices, scalable machine learning frameworks, and advanced data collection techniques using diverse voice datasets. The symposium highlighted the successful integration of AI in detecting and analyzing voice signals for various health applications, with significant advancements in standardization, privacy, and clinical validation processes.DiscussionThe symposium underscored the importance of interdisciplinary collaboration to address the technical, ethical, and clinical challenges in the field of voice biomarkers. While AI models have shown promise in analyzing voice data, challenges such as data variability, security, and scalability remain. Future efforts must focus on refining data collection standards, advancing ethical AI practices, and ensuring diverse dataset inclusion to improve model robustness. By fostering collaboration among researchers, clinicians, and technologists, the symposium laid a foundation for future innovations in AI-driven voice analysis for healthcare diagnostics and treatment.

Abstract The molecular characteristics of metastatic upper tract urothelial carcinoma (UTUC) are not well understood, and there is a lack of knowledge regarding the genomic and transcriptomic differences between primary and metastatic UTUC. To address these gaps, we integrate whole-exome sequencing, RNA sequencing, and Imaging Mass Cytometry using lanthanide metal-conjugated antibodies of 44 tumor samples from 28 patients with high-grade primary and metastatic UTUC. We perform a spatially-resolved single-cell analysis of cancer, immune, and stromal cells to understand the evolution of primary to metastatic UTUC. We discover that actionable genomic alterations are frequently discordant between primary and metastatic UTUC tumors in the same patient. In contrast, molecular subtype membership and immune depletion signature are stable across primary and matched metastatic UTUC. Molecular and immune subtypes are consistent between bulk RNA-sequencing and mass cytometry of protein markers from 340,798 single cells. Molecular subtypes at the single-cell level are highly conserved between primary and metastatic UTUC tumors within the same patient.

Abstract We previously reported the results of a randomized phase II trial (NCT02904954) in patients with early-stage non-small cell lung cancer (NSCLC) who were treated with either two preoperative cycles of the anti-PD-L1 antibody durvalumab alone or combined with immunomodulatory doses of stereotactic radiation (DRT). The trial met its primary endpoint of major pathological response, which was significantly higher following DRT with no new safety signals. Here, we report on the prespecified secondary endpoint of disease-free survival (DFS) regardless of treatment assignment and the prespecified exploratory analysis of DFS in each arm of the trial. DFS at 2 and 3 years across patients in both arms of the trial were 73% (95% CI: 62.1–84.5) and 65% (95% CI: 52.5–76.9) respectively. For the exploratory endpoint of DFS in each arm of the trial, three-year DFS was 63% (95% CI: 46.0–80.4) in the durvalumab monotherapy arm compared to 67% (95% CI: 49.6–83.4) in the dual therapy arm. In addition, we report post hoc exploratory analysis of progression-free survival as well as molecular correlates of response and recurrence through high-plex immunophenotyping of sequentially collected peripheral blood and gene expression profiles from resected tumors in both treatment arms. Together, our results contribute to the evolving landscape of neoadjuvant treatment regimens for NSCLC and identify easily measurable potential biomarkers of response and recurrence.

Prostate cancer is a heritable disease with ancestry-biased incidence and mortality. Polygenic risk scores (PRSs) offer promising advancements in predicting disease risk, including prostate cancer. While their accuracy continues to improve, research aimed at enhancing their effectiveness within African and Asian populations remains key for equitable use. Recent algorithmic developments for PRS derivation have resulted in improved pan-ancestral risk prediction for several diseases. In this study, we benchmark the predictive power of six widely used PRS derivation algorithms, including four of which adjust for ancestry, against prostate cancer cases and controls from the UK Biobank and All of Us cohorts. We find modest improvement in discriminatory ability when compared with a simple method that prioritizes variants, clumping, and published polygenic risk scores. Our findings underscore the importance of improving upon risk prediction algorithms and the sampling of diverse cohorts.

Abstract Intracranial metastases in prostate cancer are uncommon but clinically aggressive. A detailed molecular characterization of prostate cancer intracranial metastases would improve our understanding of their pathogenesis and the search for new treatment strategies. We evaluated the clinical and molecular characteristics of 36 patients with metastatic prostate cancer to either the dura or brain parenchyma. We performed whole genome sequencing (WGS) of 10 intracranial prostate cancer metastases, as well as WGS of primary prostate tumors from men who later developed metastatic disease (n = 6) and nonbrain prostate cancer metastases (n = 36). This first study focused on WGS of prostate intracranial metastases led to several new insights. First, there was a higher diversity of complex structural alterations in prostate cancer intracranial metastases compared to primary tumor tissues. Chromothripsis and chromoplexy events seemed to dominate, yet there were few enrichments of specific categories of structural variants compared with non-brain metastases. Second, aberrations involving the AR gene, including AR enhancer gain were observed in 7/10 (70%) of intracranial metastases, as well as recurrent loss of function aberrations involving TP53 in 8/10 (80%), RB1 in 2/10 (20%), BRCA2 in 2/10 (20%), and activation of the PI3K/AKT/PTEN pathway in 8/10 (80%). These alterations were frequently present in tumor tissues from other sites of disease obtained concurrently or sequentially from the same individuals. Third, clonality analysis points to genomic factors and evolutionary bottlenecks that contribute to metastatic spread in patients with prostate cancer. These results describe the aggressive molecular features underlying intracranial metastasis that may inform future diagnostic and treatment approaches.

OBJECTIVES/GOALS: The goal of this project is to characterize the efficacy of FOXA2 as a potential biomarker for patients with metastatic castrate-resistant prostate adenocarcinoma (CRPC) before transitioning to neuroendocrine prostate cancer (NEPC). NEPC currently has no therapeutic options and poor mechanistic understand of its origins. METHODS/STUDY POPULATION: In our study, we have utilized a multi-omics approach to characterize the potential efficacy of FOXA2 as a prognostic biomarker in numerous patient-derived castrate-resistant prostate cancer (CRPC) models. We have performed ATAC-, ChIP-, RNA-seq and proteomics to fully characterize where FOXA2 is binding genome-wide, how FOXA2 alters chromatin accessibility dynamics, identify regulatory gene targets of FOXA2, and identify FOXA2 protein-protein interactors. We have supported these findings using publicly available data from independent CRPC and NEPC patient cohorts and prostate cancer cell models. RESULTS/ANTICIPATED RESULTS: Our findings show that FOXA2 overexpression suppressed androgen signaling and promoted progression to a NEPC phenotype under short- and long-term androgen deprivation conditions, respectively. Further, FOXA2 redirected the chromatin accessibility landscape to be consistent with an NEPC gene expression program, including increased chromatin accessibility for key NEPC transcription factors. FOXA2 ChIP-seq showed FOXA2 to be bound at known NEPC driver genes and epigenetic modifiers across multiple stages of prostate cancer progression. Lastly, we discovered that FOXA2 physically interacts with key NEPC TFs and epigenetic regulators, suggesting that these FOXA2 physical interactions are required for NEPC progression. DISCUSSION/SIGNIFICANCE: This project will determine the efficacy of FOXA2 as a biomarker in advanced prostatecancer samples, which will translate as a potentially useful tool for clinicians to use for treatment of advanced prostate cancer patients.

Purpose: Spina bifida (SB) arises from complex genetic interactions that converge to interfere with neural tube closure. Understanding the precise patterns conferring SB risk requires a deep exploration of the genomic networks and molecular pathways that govern neurulation. This study aims to delineate genome-wide regulatory signatures underlying SB pathophysiology, Methods: An untargeted, genome-wide approach was used to interrogate regulatory regions for rare single-nucleotide and copy-number variants (rSNVs and rCNVs, respectively) predicted to affect gene expression, comparing results from SB patients with healthy controls. Qualifying variants were subjected to a deep learning prioritization framework to identify the most functionally relevant variants, as well as the likely target genes affected by these rare regulatory variants. Results: This ensemble of computational tools identified rSNVs in specific transcription factor binding sites (TFBSs) that distinguish SB cases from controls. rSNV enrichment was found in specific TFBSs, especially CCCTC-binding factor binding sites. These TFBSs were subjected to a deep learning prioritization framework to identify the most functionally relevant variants, as well as the likely target genes affected by these rSNVs. The functional pathways or modules implicated by these regulated genes serve protein transport, cilia assembly, and central nervous system development. Moreover, the detected rare copy-number variants in SB cases are positioned to disrupt gene regulatory networks and alter 3-dimensional genomic architectures, including brain-specific enhancers and topologically associated domain boundaries of relevant cell types. Conclusion: Our study provides a resource for identifying and interpreting genomic regulatory DNA variant contributions to human SB genetic predisposition.

Cancer of unknown primary (CUP) represents a significant diagnostic and therapeutic challenge, being the third to fourth leading cause of cancer death, despite advances in diagnostic tools. This article presents a successful approach using a novel genomic analysis in the evaluation and treatment of a CUP patient, leveraging whole-exome sequencing (WES) and RNA sequencing (RNA-seq). The patient, with a history of multiple primary tumors including urothelial cancer, exhibited a history of rapid progression on empirical chemotherapy. The application of our approach identified a molecular target, characterized the tumor expression profile and the tumor microenvironment, and analyzed the origin of the tumor, leading to a tailored treatment. This resulted in a substantial radiological response across all metastatic sites and the predicted primary site of the tumor. We argue that a comprehensive genomic and molecular profiling approach, like the BostonGene© Tumor Portrait, can provide a more definitive, personalized treatment strategy, overcoming the limitations of current predictive assays. This approach offers a potential solution to an unmet clinical need for a standardized approach in identifying the tumor origin for the effective management of CUP.

Abstract Background Lung cancer remains the major cause of cancer‐related deaths worldwide. Early stages of lung cancer are characterized by long asymptomatic periods that are ineffectively identified with the current screening programs. This deficiency represents a lost opportunity to improve the overall survival of patients. Serum biomarkers are among the most effective strategies for cancer screening and follow up. Methods Using bead‐based multiplexing assays we screened plasma and tumours of the KrasG12D/+; Lkb1f/f (KL) mouse model of lung cancer for cytokines that could be used as biomarkers. We identified tissue inhibitor of metalloproteinase 1 (TIMP1) as an early biomarker and validated this finding in the plasma of lung cancer patients. We used immunohistochemistry (IHC), previously published single‐cell RNA‐seq and bulk RNA‐seq data to assess the source and expression of TIMP1in the tumour. The prognostic value of TIMP1 was assessed using publicly available human proteomic and transcriptomic databases. Results We found that TIMP1 is a tumour‐secreted protein with high sensitivity and specificity for aggressive cancer, even at early stages in mice. We showed that TIMP1 levels in the tumour and serum correlate with tumour burden and worse survival in mice. We validated this finding using clinical samples from our institution and publicly available human proteomic and transcriptomic databases. These data support the finding that high tumour expression of TIMP1 correlates with an unfavorable prognosis in lung cancer patients. Conclusion TIMP1 is a suitable biomarker for lung cancer detection.

The IRE1α-XBP1 arm of the unfolded protein response (UPR) has been associated with immunosuppression and cancer progression. Here the authors show that IRE1α-XBP1 activation is associated with poor overall survival in patients with non-small cell lung cancer and that IRE1α loss in cancer cells promotes anti-tumor immune responses in lung cancer preclinical models.

Abstract Spatial omics technologies enable a deeper understanding of cellular organizations and interactions within a tissue of interest. These assays can identify specific compartments or regions in a tissue with differential transcript or protein abundance, delineate their interactions, and complement other methods in defining cellular phenotypes. A variety of spatial methodologies are being developed and commercialized; however, these techniques differ in spatial resolution, multiplexing capability, scale/throughput, and coverage. Here, we review the current and prospective landscape of single cell to subcellular resolution spatial omics technologies and analysis tools to provide a comprehensive picture for both research and clinical applications.

Single-cell RNA sequencing (scRNA-seq) has become a routinely used technique to quantify the gene expression profile of thousands of single cells simultaneously. Analysis of scRNA-seq data plays an important role in the study of cell states and phenotypes, and has helped elucidate biological processes, such as those occurring during the development of complex organisms, and improved our understanding of disease states, such as cancer, diabetes, and coronavirus disease 2019 (COVID-19). Deep learning, a recent advance of artificial intelligence that has been used to address many problems involving large datasets, has also emerged as a promising tool for scRNA-seq data analysis, as it has a capacity to extract informative and compact features from noisy, heterogeneous, and high-dimensional scRNA-seq data to improve downstream analysis. The present review aims at surveying recently developed deep learning techniques in scRNA-seq data analysis, identifying key steps within the scRNA-seq data analysis pipeline that have been advanced by deep learning, and explaining the benefits of deep learning over more conventional analytic tools. Finally, we summarize the challenges in current deep learning approaches faced within scRNA-seq data and discuss potential directions for improvements in deep learning algorithms for scRNA-seq data analysis.

Primary clear cell renal cell carcinoma (ccRCC) has been previously characterized, but the genomic landscape of metastatic ccRCC is largely unexplored. Here, we performed whole exome sequencing (WES) in 68 samples from 44 patients with ccRCC, including 52 samples from a metastatic site. SETD2, PBRM1, APC and VHL were the most frequently mutated genes in the metastatic ccRCC cohort. RBM10 and FBXW7 were also among the 10 most frequently mutated genes in metastatic tissues. Recurrent somatic copy number variations (CNV) were observed at the previously identified regions 3p25, 9p21 and 14q25, but also at 6p21 (CDKN1A) and 13q14 (RB1). No statistically significant differences were found between samples from therapy‐naïve and pretreated patients. Clonal evolution analyses with multiple samples from 13 patients suggested that early appearance of CNVs at 3p25, 9p21 and 14q25 may be associated with rapid clinical progression. Overall, the genomic landscapes of primary and metastatic ccRCC seem to share frequent CNVs at 3p25, 9p21 and 14q25. Future work will clarify the implication of RBM10 and FBXW7 mutations and 6p21 and 13q14 CNVs in metastatic ccRCC.

Abstract The Global Alliance for Genomics and Health (GA4GH) is developing a suite of coordinated standards for genomics for healthcare. The Phenopacket is a new GA4GH standard for sharing disease and phenotype information that characterizes an individual person, linking that individual to detailed phenotypic descriptions, genetic information, diagnoses, and treatments. A detailed example is presented that illustrates how to use the schema to represent the clinical course of a patient with retinoblastoma, including demographic information, the clinical diagnosis, phenotypic features and clinical measurements, an examination of the extirpated tumor, therapies, and the results of genomic analysis. The Phenopacket Schema, together with other GA4GH data and technical standards, will enable data exchange and provide a foundation for the computational analysis of disease and phenotype information to improve our ability to diagnose and conduct research on all types of disorders, including cancer and rare diseases.

Summary: Recent studies have shown the importance of the dynamic tumor microenvironment (TME) in high-grade gliomas (HGGs). In particular, myeloid cells are known to mediate immunosuppression in glioma; however, it is still unclear if myeloid cells play a role in low-grade glioma (LGG) malignant progression. Here, we investigate the cellular heterogeneity of the TME using single-cell RNA sequencing in a murine glioma model that recapitulates the malignant progression of LGG to HGG. LGGs show increased infiltrating CD4+ and CD8+ T cells and natural killer (NK) cells in the TME, whereas HGGs abrogate this infiltration. Our study identifies distinct macrophage clusters in the TME that show an immune-activated phenotype in LGG but then evolve to an immunosuppressive state in HGG. We identify CD74 and macrophage migration inhibition factor (MIF) as potential targets for these distinct macrophage populations. Targeting these intra-tumoral macrophages in the LGG stage may attenuate their immunosuppressive properties and impair malignant progression.

Abstract Objective This study aims to develop and validate a convolutional neural network (CNN)‐based algorithm for automatic selection of informative frames in flexible laryngoscopic videos. The classifier has the potential to aid in the development of computer‐aided diagnosis systems and reduce data processing time for clinician‐computer scientist teams. Methods A dataset of 22,132 laryngoscopic frames was extracted from 137 flexible laryngostroboscopic videos from 115 patients. 55 videos were from healthy patients with no laryngeal pathology and 82 videos were from patients with vocal fold polyps. The extracted frames were manually labeled as informative or uninformative by two independent reviewers based on vocal fold visibility, lighting, focus, and camera distance, resulting in 18,114 informative frames and 4018 uninformative frames. The dataset was split into training and test sets. A pre‐trained ResNet‐18 model was trained using transfer learning to classify frames as informative or uninformative. Hyperparameters were set using cross‐validation. The primary outcome was precision for the informative class and secondary outcomes were precision, recall, and F1‐score for all classes. The processing rate for frames between the model and a human annotator were compared. Results The automated classifier achieved an informative frame precision, recall, and F1‐score of 94.4%, 90.2%, and 92.3%, respectively, when evaluated on a hold‐out test set of 4438 frames. The model processed frames 16 times faster than a human annotator. Conclusion The CNN‐based classifier demonstrates high precision for classifying informative frames in flexible laryngostroboscopic videos. This model has the potential to aid researchers with dataset creation for computer‐aided diagnosis systems by automatically extracting relevant frames from laryngoscopic videos.

Clinical performance status is designed to be a measure of overall health, reflecting a patient's physiological reserve and ability to tolerate various forms of therapy. Currently, it is measured by a combination of subjective clinician assessment and patient-reported exercise tolerance in the context of daily living activities. In this study, we assess the feasibility of combining objective data sources and patient-generated health data (PGHD) to improve the accuracy of performance status assessment during routine cancer care. Patients undergoing routine chemotherapy for solid tumors, routine chemotherapy for hematologic malignancies, or hematopoietic stem cell transplant (HCT) at one of four sites in a cancer clinical trials cooperative group were consented to a six-week prospective observational clinical trial (NCT02786628). Baseline data acquisition included cardiopulmonary exercise testing (CPET) and a six-minute walk test (6MWT). Weekly PGHD included patient-reported physical function and symptom burden. Continuous data capture included use of a Fitbit Charge HR (sensor). Baseline CPET and 6MWT could only be obtained in 68% of study patients, suggesting low feasibility during routine cancer treatment. In contrast, 84% of patients had usable fitness tracker data, 93% completed baseline patient-reported surveys, and overall, 73% of patients had overlapping sensor and survey data that could be used for modeling. A linear model with repeated measures was constructed to predict the patient-reported physical function. Sensor-derived daily activity, sensor-derived median heart rate, and patient-reported symptom burden emerged as strong predictors of physical function (marginal R2 0.429-0.433, conditional R2 0.816-0.822). Trial Registration: Clinicaltrials.gov Id NCT02786628.

Targeting lineage-defined transcriptional dependencies has emerged as an effective therapeutic strategy in cancer treatment. Through screening for molecular vulnerabilities of mantle cell lymphoma (MCL), we identified a set of transcription factors (TFs) including FOXO1, EBF1, PAX5, and IRF4 that are essential for MCL propagation. Integrated chromatin immunoprecipitation and sequencing (ChIP-Seq) with transcriptional network reconstruction analysis revealed FOXO1 as a master regulator that acts upstream in the regulatory TF hierarchy. FOXO1 is both necessary and sufficient to drive MCL lineage commitment through supporting the lineage-specific transcription programs. We further show that FOXO1, but not its close paralog FOXO3, can reprogram myeloid leukemia cells and induce B-lineage gene expression. Finally, we demonstrate that cpd10, a small molecule identified from an enriched FOXO1 inhibitor library, induces a robust cytotoxic response in MCL cells in vitro and suppresses MCL progression in vivo. Our findings establish FOXO1 inhibition as a therapeutic strategy targeting lineage-driven transcriptional addiction in MCL.

Pancreatic neuroendocrine tumors (PNETs) are malignancies arising from the islets of Langerhans. Therapeutic options are limited for the over 50% of patients who present with metastatic disease. We aimed to identify mechanisms to remodel the PNET tumor microenvironment (TME) to ultimately enhance susceptibility to immunotherapy. The TMEs of localized and metastatic PNETs were investigated using an approach that combines RNA-Seq, cancer and T cell profiling, and pharmacologic perturbations. RNA-Seq analysis indicated that the primary tumors of metastatic PNETs showed significant activation of inflammatory and immune-related pathways. We determined that metastatic PNETs featured increased numbers of tumor-infiltrating T cells compared with localized tumors. T cells isolated from both localized and metastatic PNETs showed evidence of recruitment and antigen-dependent activation, suggestive of an immune-permissive microenvironment. A computational analysis suggested that vorinostat, a histone deacetylase inhibitor, may perturb the transcriptomic signature of metastatic PNETs. Treatment of PNET cell lines with vorinostat increased chemokine CCR5 expression by NF-κB activation. Vorinostat treatment of patient-derived metastatic PNET tissues augmented recruitment of autologous T cells, and this augmentation was substantiated in a mouse model of PNET. Pharmacologic induction of chemokine expression may represent a promising approach for enhancing the immunogenicity of metastatic PNET TMEs.

Abstract Nicotinamide (NAM, a variant of vitamin B3) has recently been shown to accelerate the activation of human CD4+ and CD8+ T cells exposed to repeated CD3/CD28 agonism in vitro. Here, we demonstrate that T cells infiltrating mouse mammary carcinomas that are therapeutically controlled by NAM also express multiple markers of late-stage activation. Taken together, these findings lend additional support to the notion that the antineoplastic effects of NAM involve at least some degree of restored cancer immunosurveillance.

Abstract Both clinical and genetic factors drive the risk of venous thromboembolism. However, whether clinically recorded risk factors and genetic variants can be combined into a clinically applicable predictive score remains unknown. Using Cox proportional-hazard models, we analyzed the association of risk factors with the likelihood of venous thromboembolism in U.K. Biobank, a large prospective cohort. We then created a polygenic risk score of 36 single nucleotide polymorphisms and a clinical score determined by age, sex, body mass index, previous cancer diagnosis, smoking status, and fracture in the last 5 years. Participants were at significantly increased risk of venous thromboembolism if they were at high clinical risk (subhazard ratio, 4.37 [95% CI, 3.85–4.97]) or high genetic risk (subhazard ratio, 3.02 [95% CI, 2.63–3.47]) relative to participants at low clinical or genetic risk, respectively. The combined model, consisting of clinical and genetic components, was significantly better than either the clinical or the genetic model alone (P

Chronic lymphocytic leukemia has been studied using multiple levels of omics data. Here, the authors use exome sequencing, SNP, protein and gene expression data to identify distinct biologic tumor subtypes with heterogeneous prognostic impact after chemo- or immunochemotherapy.

Although immune checkpoint inhibitors (ICIs) are increasingly used as second-line treatments for urothelial cancer (UC), only a small proportion of patients respond. Therefore, understanding the mechanisms of response to ICIs is critical to improve clinical outcomes for UC patients. The tumor microenvironment (TME) is recognized as a key player in tumor progression and the response to certain anti-cancer treatments. This study aims to investigate the mechanism of response using integrated genomic and transcriptomic profiling of a UC patient who was part of the KEYNOTE-045 trial and showed an exceptional response to pembrolizumab. Diagnosed in 2014 and receiving first-line chemotherapy without success, the patient took part in the KEYNOTE-045 trial for 2 years. She showed dramatic improvement and has now been free of disease for over 6 years. Recently described by Bagaev et al., the Molecular Functional (MF) Portrait was utilized to dissect genomic and transcriptomic features of the patient’s tumor and TME. The patient’s tumor was characterized as Immune Desert, which is suggestive of a non-inflamed microenvironment. Integrated whole-exome sequencing (WES) and RNA sequencing (RNA-seq) analysis identified an ATM mutation and high TMB level (33.9 mut/mb), which are both positive biomarkers for ICI response. Analysis further revealed the presence of the APOBEC complex, indicating the potential for use of APOBEC signatures as predictive biomarkers for immunotherapy response. Overall, comprehensive characterization of the patient’s tumor and TME with the MF Portrait revealed important insights that could potentially be hypothesis generating to identify clinically useful biomarkers and improve treatment for UC patients.

Abstract Background Interest in the application of machine learning (ML) to the design, conduct, and analysis of clinical trials has grown, but the evidence base for such applications has not been surveyed. This manuscript reviews the proceedings of a multi-stakeholder conference to discuss the current and future state of ML for clinical research. Key areas of clinical trial methodology in which ML holds particular promise and priority areas for further investigation are presented alongside a narrative review of evidence supporting the use of ML across the clinical trial spectrum. Results Conference attendees included stakeholders, such as biomedical and ML researchers, representatives from the US Food and Drug Administration (FDA), artificial intelligence technology and data analytics companies, non-profit organizations, patient advocacy groups, and pharmaceutical companies. ML contributions to clinical research were highlighted in the pre-trial phase, cohort selection and participant management, and data collection and analysis. A particular focus was paid to the operational and philosophical barriers to ML in clinical research. Peer-reviewed evidence was noted to be lacking in several areas. Conclusions ML holds great promise for improving the efficiency and quality of clinical research, but substantial barriers remain, the surmounting of which will require addressing significant gaps in evidence.

Abstract The availability of fresh frozen (FF) tissue is a barrier for implementing RNA sequencing (RNA-seq) in the clinic. The majority of clinical samples are stored as formalin-fixed, paraffin-embedded (FFPE) tissues. Exome capture platforms have been developed for RNA-seq from FFPE samples. However, these methods have not been systematically compared. We performed transcriptomic analysis of 32 FFPE tumor samples from 11 patients using three exome capture-based methods: Agilent SureSelect V6, TWIST NGS Exome, and IDT XGen Exome Research Panel. We compared these methods to the TruSeq RNA-seq of fresh frozen (FF-TruSeq) tumor samples from the same patients. We assessed the recovery of clinically relevant biological features. The Spearman’s correlation coefficients between the global expression profiles of the three capture-based methods from FFPE and matched FF-TruSeq were high (rho = 0.72–0.9, p

The heterogeneity of tumor evolution from AR-positive, adenocarcinoma to AR-negative, neuroendocrine prostate cancer (NEPC) is not fully characterized. Here the authors generate a mouse model to show that Rb1 loss and MYCN overexpression accelerates the progression to AR-negative NEPC and identify emergence of distinct subpopulations of NEPC cells.

Summary: Background: Estimating tumor purity is especially important in the age of precision medicine. Purity estimates have been shown to be critical for correction of tumor sequencing results, and higher purity samples allow for more accurate interpretations from next-generation sequencing results. Molecular-based purity estimates using computational approaches require sequencing of tumors, which is both time-consuming and expensive. Methods: Here we propose an approach, weakly-supervised purity (wsPurity), which can accurately quantify tumor purity within a digitally captured hematoxylin and eosin (H&E) stained histological slide, using several types of cancer from The Cancer Genome Atlas (TCGA) as a proof-of-concept. Findings: Our model predicts cancer type with high accuracy on unseen cancer slides from TCGA and shows promising generalizability to unseen data from an external cohort (F1-score of 0.83 for prostate adenocarcinoma). In addition we compare performance of our model on tumor purity prediction with a comparable fully-supervised approach on our TCGA held-out cohort and show our model has improved performance, as well as generalizability to unseen frozen slides (0.1543 MAE on an independent test cohort). In addition to tumor purity prediction, our approach identified high resolution tumor regions within a slide, and can also be used to stratify tumors into high and low tumor purity, using different cancer-dependent thresholds. Interpretation: Overall, we demonstrate our deep learning model's different capabilities to analyze tumor H&E sections. We show our model is generalizable to unseen H&E stained slides from data from TCGA as well as data processed at Weill Cornell Medicine. Funding: Starr Cancer Consortium Grant (SCC I15-0027) to Iman Hajirasouliha.

Abstract The epichaperome is a new cancer target composed of hyperconnected networks of chaperome members that facilitate cell survival. Cancers with an altered chaperone configuration may be susceptible to epichaperome inhibitors. We developed a flow cytometry-based assay for evaluation and monitoring of epichaperome abundance at the single cell level, with the goal of prospectively identifying patients likely to respond to epichaperome inhibitors, to measure target engagement, and dependency during treatment. As proof of principle, we describe a patient with an unclassified myeloproliferative neoplasm harboring a novel PML-SYK fusion, who progressed to acute myeloid leukemia despite chemotherapy and allogeneic stem cell transplant. The leukemia was identified as having high epichaperome abundance. We obtained compassionate access to an investigational epichaperome inhibitor, PU-H71. After 16 doses, the patient achieved durable complete remission. These encouraging results suggest that further investigation of epichaperome inhibitors in patients with abundant baseline epichaperome levels is warranted.

OBJECTIVES/GOALS: Anti-cancer therapies, such as chemotherapy, can induce senescence. Senescent cells may produce factors that can promote tumor progression. In this study, we will investigate the effect of senolytics and anti-cancer treatment on fibroblasts, which are a part of the tumor microenvironment, and patient-derived colorectal cancer organoids. METHODS/STUDY POPULATION: We will induce senescence in fibroblast lines via irradation. Induction of senescence will be confirmed by monitoring SASP production, changes in morphology and proliferation rates, and senescence-associated b-galactosidase activity. To investigate the efficacy of senolytics on senescence-induced fibroblasts and CRC tumor organoids, we will creat a dose response curve and calculate IC50 values for proliferating fibroblast, senescent fibroblasts and CRC organoids. To identify the synergistic effects of anti-cancer and senolytic compounds, including Navitoclax and Dasatinib, on fibroblasts and CRC organoids, we will create dose matrixes using senolytics at concentrations that were shown to have senolytic activity and drugs from an anti-cancer library. RESULTS/ANTICIPATED RESULTS: If senescence is induced in the fibroblast lines, we expect to see no changes in confluency over 4 days, the morphology will change from a thin, spindly shape to a flattened shape, and senescence-associated b-galactosidase activity will be observed. After the fibroblast lines are treated with potential senolytic compounds, we would expect to see decreased viability in the senescence-induced fibroblast lines when compared to proliferating fibroblast lines. We predict that the viability of CRC organoid lines will slightly decrease at high concentrations of the senolytic due to overall toxicity. We expect that the senolytic and anti-cancer compounds will have a synergistic effect. Senolytic activity could reduce the senescent cell population that was developed in response to anti-cancer therapy. DISCUSSION/SIGNIFICANCE: There is an increased interest in identifying compounds that selectively promote apoptosis in senescent cells. This study uses a cell-based approach to validate senolytic activity of compounds with senolytic potential in senescence-induced fibroblast lines and investigates the synergistic effects of senolytics and anti-cancer compounds on CRC.

The role of germline variation in human cancers is not fully understood. Here, the authors define the landscape of putative deleterious germline variants that abrogate tumor suppressor proteins in advanced urothelial cancer patients.

Summary: The effect of SARS-CoV-2 infection on placental function is not well understood. Analysis of placentas from women who tested positive at delivery showed SARS-CoV-2 genomic and subgenomic RNA in 22 out of 52 placentas. Placentas from two mothers with symptomatic COVID-19 whose pregnancies resulted in adverse outcomes for the fetuses contained high levels of viral Alpha variant RNA. The RNA was localized to the trophoblasts that cover the fetal chorionic villi in direct contact with maternal blood. The intervillous spaces and villi were infiltrated with maternal macrophages and T cells. Transcriptome analysis showed an increased expression of chemokines and pathways associated with viral infection and inflammation. Infection of placental cultures with live SARS-CoV-2 and spike protein-pseudotyped lentivirus showed infection of syncytiotrophoblast and, in rare cases, endothelial cells mediated by ACE2 and Neuropilin-1. Viruses with Alpha, Beta, and Delta variant spikes infected the placental cultures at significantly greater levels.

Summary: To investigate changes in the tumor microenvironment (TME) during lung cancer progression, we interrogate tumors from two chest computed tomography (CT)-defined groups. Pure non-solid (pNS) CT density nodules contain preinvasive/minimally invasive cancers, and solid density nodules contain invasive cancers. Profiling data reveal a dynamic interaction between the tumor and its TME throughout progression. Alterations in genes regulating the extracellular matrix and genes regulating fibroblasts are central at the preinvasive state. T cell-mediated immune suppression is initiated in preinvasive nodules and sustained with rising intensity through progression to invasive tumors. Reduced T cell infiltration of the cancer cell nests is more frequently associated with preinvasive cancers, possibly until tumor evolution leads to a durable, viable invasive phenotype accompanied by more varied and robust immune suppression. Upregulation of immune checkpoints occurs only in the invasive nodules. Throughout progression, an effector immune response is present but is effectively thwarted by the immune-suppressive elements.

Current preclinical models to investigate human HR + breast cancer progression and response to immunotherapy in vivo are limited. Here, the authors demonstrate that mammary tumours driven by a synthetic progestin combined with an oral carcinogen recapitulate several immunobiological features of human HR + breast cancers.

Abstract Single-cell RNA-sequencing (scRNA-seq) facilitates the unbiased reconstruction of multicellular tissue systems in health and disease. Here, we present a curated scRNA-seq dataset of human muscle samples from 10 adult donors with diverse anatomical locations. We integrated ~ 22,000 single-cell transcriptomes using Scanorama to account for technical and biological variation and resolved 16 distinct populations of muscle-resident cells using unsupervised clustering of the data compendium. These cell populations included muscle stem/progenitor cells (MuSCs), which bifurcated into discrete “quiescent” and “early-activated” MuSC subpopulations. Differential expression analysis identified transcriptional profiles altered in the activated MuSCs including genes associated with aging, obesity, diabetes, and impaired muscle regeneration, as well as long non-coding RNAs previously undescribed in human myogenic cells. Further, we modeled ligand-receptor cell-communication interactions and observed enrichment of the TWEAK-FN14 pathway in activated MuSCs, a characteristic signature of muscle wasting diseases. In contrast, the quiescent MuSCs have enhanced expression of the EGFR receptor, a recognized human MuSC marker. This work provides a new benchmark reference resource to examine human muscle tissue heterogeneity and identify potential targets in MuSC diversity and dysregulation in disease contexts.

Abstract BCOR has been recognized as a recurrently altered gene in a subset of pediatric tumors of the central nervous system (CNS). Here, we describe a novel BCOR-CREBBP fusion event in a case of pediatric infiltrating astrocytoma and further probe the frequency of related fusion events in CNS tumors. We analyzed biopsy samples taken from a 15-year-old male with an aggressive, unresectable and multifocal infiltrating astrocytoma. We performed RNA sequencing (RNA-seq) and targeted DNA sequencing. In the index case, the fused BCOR-CREBBP transcript comprises exons 1–4 of BCOR and exon 31 of CREBBP. The fused gene thus retains the Bcl6 interaction domain of BCOR while eliminating the domain that has been shown to interact with the polycomb group protein PCGF1. The fusion event was validated by FISH and reverse transcriptase PCR. An additional set of 177 pediatric and adult primary CNS tumors were assessed via FISH for BCOR break apart events, all of which were negative. An additional 509 adult lower grade infiltrating gliomas from the publicly available TCGA dataset were screened for BCOR or CREBBP fusions. In this set, one case was found to harbor a CREBBP-GOLGA6L2 fusion and one case a CREBBP-SRRM2 fusion. In a third patient, both BCOR-L3MBTL2 and EP300-BCOR fusions were seen. Of particular interest to this study, EP300 is a paralog of CREBBP and the breakpoint seen involves a similar region of the gene to that of the index case; however, the resultant transcript is predicted to be completely distinct. While this gene fusion may play an oncogenic role through the loss of tumor suppressor functions of BCOR and CREBBP, further screening over larger cohorts and functional validation is needed to determine the degree to which this or similar fusions are recurrent and to elucidate their oncogenic potential.

Immune infiltration provides critical information for health and disease, yet it is unclear what factors influence infiltration levels. Here, the authors analyze human tissue transcriptomes from the Genotype-Tissue Expression project to find infiltration patterns regulated by age, sex and host genetic information.

Recent studies have utilized bulk tumour mRNA sequencing to classify bladder cancers into distinct subgroups. Here, the authors use single cell transcriptomic analysis and cell transplant studies to show that epithelial plasticity can generate basal, luminal and mesenchymal phenotypes in human and murine bladder cancers.

Benign prostatic hyperplasia (BPH) is one of the most common diseases affecting aging men with limited therapeutic options. In this study, the authors describe the molecular characterization of BPH performing genomic, transcriptomic and epigenetic analysis of 18 BPH cases.

BACKGROUND Prostate cancer is multifocal with distinct molecular subtypes. The utility of genomic subtyping has been challenged due to inter- and intrafocal heterogeneity. We sought to characterize the subtype-defining molecular alterations of primary prostate cancer across all tumor foci within radical prostatectomy (RP) specimens and determine the prevalence of collision tumors.METHODS From the Early Detection Research Network cohort, we identified 333 prospectively collected RPs from 2010 to 2014 and assessed ETS-related gene (ERG), serine peptidase inhibitor Kazal type 1 (SPINK1), phosphatase and tensin homolog (PTEN), and speckle type BTB/POZ protein (SPOP) molecular status. We utilized dual ERG/SPINK1 immunohistochemistry and fluorescence in situ hybridization to confirm ERG rearrangements and characterize PTEN deletion, as well as high-resolution melting curve analysis and Sanger sequencing to determine SPOP mutation status.RESULTS Based on index focus alone, ERG, SPINK1, PTEN, and SPOP alterations were identified in 47.5%, 10.8%, 14.3%, and 5.1% of RP specimens, respectively. In 233 multifocal RPs with ERG/SPINK1 status in all foci, 139 (59.7%) had discordant molecular alterations between foci. Collision tumors, as defined by discrepant ERG/SPINK1 status within a single focus, were identified in 29 (9.4%) RP specimens.CONCLUSION Interfocal molecular heterogeneity was identified in about 60% of multifocal RP specimens, and collision tumors were present in about 10%. We present this phenomenon as a model for the intrafocal heterogeneity observed in previous studies and propose that future genomic studies screen for collision tumors to better characterize molecular heterogeneity.FUNDING Early Detection Research Network US National Cancer Institute (NCI) 5U01 CA111275-09, Center for Translational Pathology at Weill Cornell Medicine (WCM) Department of Pathology and Laboratory Medicine, US NCI (WCM SPORE in Prostate Cancer, P50CA211024-01), R37CA215040, Damon Runyon Cancer Research Foundation, US MetLife Foundation Family Clinical Investigator Award, Norwegian Cancer Society (grant 208197), and South-Eastern Norway Regional Health Authority (grant 2019016 and 2020063).

Perturbations to the epigenome are known drivers of tumorigenesis. In melanoma, alterations in histone methyltransferases that catalyze methylation at histone 3 lysine 9 and histone 3 lysine 27—two sites of critical post-translational modification—have been reported. To study the function of these methyltransferases in melanoma, we engineered melanocytes to express histone 3 lysine-to-methionine mutations at lysine 9 and lysine 27, which are known to inhibit the activity of histone methyltransferases, in a zebrafish melanoma model. Using this system, we found that loss of histone 3 lysine 9 methylation dramatically suppressed melanoma formation and that inhibition of histone 3 lysine 9 methyltransferases in human melanoma cells increased innate immune response signatures. In contrast, loss of histone 3 lysine 27 methylation significantly accelerated melanoma formation. We identified FOXD1 as a top target of PRC2 that is silenced in melanocytes and found that aberrant overexpression of FOXD1 accelerated melanoma onset. Collectively, these data demonstrate how histone 3 lysine-to-methionine mutations can be used to uncover critical roles for methyltransferases.

Deep understanding of the SARS-CoV-2 effects on host molecular pathways is paramount for the discovery of early biomarkers of outcome of coronavirus disease 2019 (COVID-19) and the identification of novel therapeutic targets. In that light, we generated metabolomic data from COVID-19 patient blood using high-throughput targeted nuclear magnetic resonance (NMR) spectroscopy and high-dimensional flow cytometry. We find considerable changes in serum metabolome composition of COVID-19 patients associated with disease severity, and response to tocilizumab treatment. We built a clinically annotated, biologically-interpretable space for precise time-resolved disease monitoring and characterize the temporal dynamics of metabolomic change along the clinical course of COVID-19 patients and in response to therapy. Finally, we leverage joint immuno-metabolic measurements to provide a novel approach for patient stratification and early prediction of severe disease. Our results show that high-dimensional metabolomic and joint immune-metabolic readouts provide rich information content for elucidation of the host’s response to infection and empower discovery of novel metabolic-driven therapies, as well as precise and efficient clinical action.

The cohesin complex plays critical roles in genomic stability and gene expression through effects on 3D architecture. Cohesin core subunit genes are mutated across a wide cross-section of cancers, but not in germinal center (GC) derived lymphomas. In spite of this, haploinsufficiency of cohesin ATPase subunit Smc3 was shown to contribute to malignant transformation of GC B-cells in mice. Herein we explored potential mechanisms and clinical relevance of Smc3 deficiency in GC lymphomagenesis. Transcriptional profiling of Smc3 haploinsufficient murine lymphomas revealed downregulation of genes repressed by loss of epigenetic tumor suppressors Tet2 and Kmt2d. Profiling 3D chromosomal interactions in lymphomas revealed impaired enhancer-promoter interactions affecting genes like Tet2, which was aberrantly downregulated in Smc3 deficient lymphomas. Tet2 plays important roles in B-cell exit from the GC reaction, and single cell RNA-seq profiles and phenotypic trajectory analysis in Smc3 mutant mice revealed a specific defect in commitment to the final steps of plasma cell differentiation. Although Smc3 deficiency resulted in structural abnormalities in GC B-cells, there was no increase of somatic mutations or structural variants in Smc3 haploinsufficient lymphomas, suggesting that cohesin deficiency largely induces lymphomas through disruption of enhancer-promoter interactions of terminal differentiation and tumor suppressor genes. Strikingly, the presence of the Smc3 haploinsufficient GC B-cell transcriptional signature in human patients with GC-derived diffuse large B-cell lymphoma (DLBCL) was linked to inferior clinical outcome and low expression of cohesin core subunits. Reciprocally, reduced expression of cohesin subunits was an independent risk factor for worse survival int DLBCL patient cohorts. Collectively, the data suggest that Smc3 functions as a bona fide tumor suppressor for lymphomas through non-genetic mechanisms, and drives disease by disrupting the commitment of GC B-cells to the plasma cell fate.