Sloan-Kettering Inst Can Research

NIH Research Projects · FY 2026 · 2024-12

PROJECT SUMMARY Lung adenocarcinoma (LUAD) is a significant public health burden in the United States, causing an estimated 55,000 deaths annually. Despite advances in immunotherapies, chemotherapies, and oncoprotein-targeted therapies, most patients ultimately relapse and cures remain elusive. Within a tumor, LUAD cancer cells adopt a spectrum of cell states endowed with variable potential for malignant progression and treatment resistance. Whether these different cell states have different metabolic preferences, and whether targeting these preferences can eliminate specific cell states, remains poorly understood. Using cell lines derived from mouse and human LUAD tumors, we discovered that a central metabolic pathway, the tricarboxylic acid (TCA) cycle, adopts a non-canonical configuration in de-differentiated LUAD states. Manipulating non-canonical or canonical TCA cycle configurations induced differentiation or de-differentiation, respectively. Our findings raise the possibility that metabolic remodeling is required for the emergence and maintenance of cell states that drive LUAD progression and treatment resistance. The goal of the proposed work is to determine how specific TCA configurations support malignant LUAD states. We hypothesize that differentiated LUAD cell states require the canonical TCA cycle whereas de-differentiated LUAD states require the non- canonical TCA cycle. To test this hypothesis, we will manipulate components of each TCA cycle in autochthonous mouse models and human patient-derived xenograft models and monitor tumor burden, progression, and emergence of specific cell states (Aim 1); determine the TCA cycle outputs that support malignant states (Aim 2), and assess the impact of TCA cycle manipulation on resistance to targeted therapy and chemotherapies (Aim 3). The proposed experiments will reveal how central metabolic networks contribute to LUAD heterogeneity and identify pathways that can be targeted to eliminate treatment-resistant cells and improve LUAD therapy.

NIH Research Projects · FY 2026 · 2024-12

Project Summary Groundbreaking immunotherapy (IO) drugs offer durable response and improved survival in cancer patients who previously had limited treatment options. The semi-quantitative assessment of PD-L1 protein expression on tumor and/or immune cells (lymphocytes, macrophage) by a certified pathologist (“positive”, “negative”, “low”, “medium”, “high”) is the most prevalent biomarker for guiding clinical decision-making in IO. However, PD-L1 expression is difficult to score on standard immunohistochemistry (IHC) slides. The disagreement among pathologists for immune cell PD-L1 scoring is greater than 50%, which can lead to a high percentage of patients receiving IO when they are unlikely to benefit. This discordance could explain why most cancer patients do not benefit from these groundbreaking yet expensive Ios (costs > $200K / year per patient). Thus, there is an urgent unmet clinical need to identify patients who will not benefit from IO. As opposed to standard single-plex IHC, multiplex immunofluorescence (mpIF) staining, though expensive, provides the opportunity to examine panels of several markers (including tumor- and immune-specific markers) individually or simultaneously as a composite while permitting stain standardization, objective scoring, and cut-offs for all the markers; mpIF also has higher sensitivity and diagnostic prediction accuracy than IHC PD-L1 scoring. This opens up unique opportunities to leverage mpIF-stained images (with objective ground truth tumor/immune cell annotations and absolute PD-L1 intensities) coupled with recent deep learning methods to improve the explainability and interpretability of the conventional IHCs broadly. In previous work, restained and co-registered IHC/mpIF whole-slide images (WSIs) were used to create a deep learning virtual mpIF restaining algorithm, DeepLIIF (Deep Learning Inferred IF), for scoring IHC Ki67 and other nuclear markers. DeepLIIF is the only IHC scoring globally available as a public/free cloud-native platform with a user-friendly web interface and AI-ready IHC/mpIF datasets; it has been extensively validated in low- and high-resource settings. Recently, DeepLIIF was extended for more reproducible and accurate visual IHC PD-L1 scoring in tumor cells for lung cancer. The work proposed under this NIH R01 will further improve DeepLIIF PD-L1 scoring by incorporating mpIF immune cell (lymphocytes and macrophage) markers, whole-cell (rather than simple nuclear) segmentation, and large/diverse datasets across lung and bladder cancers. Additionally, the team will (1) validate DeepLIIF PD-L1 tumor and immune cell scoring on thousands of IHC PD-L1 whole-slide images (with manual readouts and ground truth IHC/mpIF for a subset) spanning different antibodies, platforms, and scoring systems, and (2) validate DeepLIIF-derived spatial biomarkers and PD-L1 scores on lung and bladder cancer datasets with clinical outcomes. Successful validation will establish DeepLIIF as an interpretable, tissue non-destructive, and cost-efficient solution to accurate IO patient stratification.

NIH Research Projects · FY 2026 · 2024-12

Abstract The maintenance of genome integrity in homologous recombination deficient (HRD) cancers is linked to the backup repair of DNA double-strand breaks (DSB). Because endogenous DSB in tumors are associated with DNA replication and rapid cell proliferation, there is prompt resection of any DSB to produce a 3’-single-strand tail, which prevents non-homologous end-joining from binding to the ends. The principal backup repair pathways are microhomology mediated end-joining or single stranded annealing in this setting. RAD52 has been identified by our laboratory to be critical for the survival of cells with a defective BRCA1-BRCA2 pathway (HRD). When homologous recombination (HR) becomes inactivated in human cancers, these alternative repair pathways are used, producing the pattern of genome wide changes seen in these cancers. In this setting, RAD52 is a key protein for DSB repair, which works either by the annealing of single-strand tails of broken DNA or by RAD51- mediated strand exchange and gene conversion. The proposed work will determine the role of RAD52, HELQ and POLQ in backup DSB repair. Additional proteins engaged in these functions will be discovered. Their relative contribution to the different types of DSB repair and to the survival of HR-defective cells will be evaluated. The consequences of this work will help to define new approaches to defining targets in the treatment of HRD cancers and allow the development of new functional assays for backup DSB repair, with the ultimate goal of defining the next generation of DNA repair inhibitors beyond PARP-inhibitors. PHS 398/2590 (Rev. 06/09) Page 1 Continuation Format Page

NIH Research Projects · FY 2025 · 2024-09

Proximity Copy Paste: a methodology for single-molecule analysis of chromosome structure. Abstract Genomics assays that report protein occupancy or the 3D arrangement of crosslinked and fragmented chromosomes are often used to understand how genomic information is accessed, copied, or repaired. While powerful, such methods only interrogate a small fraction of the available information in a sample. Important questions related to whether events are coincident or mutually exclusive; or the nature of cause-effect relationships are difficult or impossible to assay. This proposal details the development of novel approaches to map the folding and protein occupancy of entire chromosomes, with single-nucleosome resolution and single-molecule precision. Our method employs a proximity labelling approach that can uniquely and indelibly tag DNA, protein or other molecules that associate in 3D space. Importantly, the method can easily be tuned to different resolutions, to allow quantitative measurements of proximity over a range of distances.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY At the turn of the millennium, the cost of sequencing one megabase of DNA exceeded 10 million dollars. Today, decoding that same megabase costs less than a penny. This dramatic reduction in the cost of sequencing has catalyzed a genomics revolution and resulted in petabases (millions of billions of bases) of DNA and RNA- sequences from human cells and tissues. We hypothesize that this transformational capacity of next-generation sequencing will now catalyze our understanding of the human virome—both by illuminating the set of DNA viruses in human tissues and by profiling the host (human) cells that are infected. Here, we outline new computational and experimental methods to realize the unique potential of petabase-scale sequencing data in studying human DNA viruses. Our work will uncover foundational aspects of the human virome, including tissue tropism and cellular reservoirs for all DNA viruses. Further, we present complementary strategies through host genetics analyses and single-cell multi-omics to define and characterize the molecular interactions of human cells associated with viral infections and latency. First, in Aim 1, we will develop methods to quantify latent viral features from petabases of unmapped whole genome sequencing reads from hundreds of thousands of individuals. These new molecular variables will reveal the degree of latent viral DNA in blood and will be paired with comprehensive host genotyping and phenotyping. We will determine host genetics factors associated with high viral levels and nominate phenotypes, including complex disease, that may be driven by long-term latent infection in individuals. In Aim 2, we will extend our petabase-scale resource (Serratus) to create a ‘Digital Human Virome’ by uniformly processing billions of dollars of public sequencing data from human cells and tissues to identify and quantify all DNA viruses. Our resource will aggregate meta-data to extract sex, cell/tissue of origin, disease status, and geographic location to create a Digital Human Virome for DNA viruses, revealing tissue tropism that can be mined for clinical associations, such as our recent discovery of HHV-6 reactivation in CAR T cells. Finally, in Aim 3, we will develop a new high-throughput single-cell multi-omics technology termed ‘Latent-seq’ that will identify individual human cells that harbor latent viruses with paired high-quality cell state measurements. We will first establish and benchmark the assay using a set of well-defined cell lines before extending applications to primary human tissues in collaboration with the Human Virome Program Consortium. Together, these workflows will define human DNA virome in health and disease by leveraging this unique moment in the capacity of genomics technologies. As every human gets infected by endemic eukaryotic DNA viruses, but only some individuals ever show symptoms, our systematic approaches will uncover new associations between molecular interactions and human phenotypes intertwined with the human virome.

NIH Research Projects · FY 2025 · 2024-09

ABSTRACT Gastric cancer is one of the most common and deadly cancers globally, accounting for over one million new cancer diagnoses and 783,000 deaths in 2018. Infection of Helicobacter pylori (H. pylori) is the most important risk factor and a necessary cause for non-cardia gastric cancer. While targeting high-risk populations who are infected with H. pylori seems a compelling strategy for gastric cancer screening and prevention, a significant concern not addressed is that only ~1 to 3% of the H. pylori infected individuals develop gastric cancer, indicating that a strategy based on H. pylori status alone is insufficient, particularly among populations with a relatively low prevalence of H. pylori infection. Thus, discovering and validating additional biomarkers, especially non-invasive ones, is necessary to facilitate the development of novel risk assessment tools for gastric cancer. The human metabolome reflects endogenous responses to exposures, including environmental and lifestyle factors that are related to carcinogenesis. The use of advanced metabolomics techniques in prospective population -based cohort studies has successfully identified both novel etiologic factors and risk biomarkers for different cancers. However, few prior studies have focused on identifying novel risk biomarkers for gastric cancer using metabolomics techniques, particularly in prospective design settings. We recently conducted the first nested case-control study within two large prospective cohorts. Our study identified 5 metabolites associated with risk of gastric cancer after correction for multiple comparisons, and 13 additional metabolite candidates for further investigations. These include metabolites related to vitamin B12 deficiency, green tea/coffee intake, and two progestin steroids, representing a few biologically plausible mechanisms involved in gastric cancer etiology. Herein, we propose to conduct a large prospective investigation within a newly formed international consortium to validate these highly promising findings, as well as discovering novel risk biomarkers, using both untargeted and targeted metabolomics approaches. Pre-diagnostic plasma samples and epidemiologic data including demographics, lifestyle factors, and dietary intakes will be harmonized for ~1,200 incident cases and 1,600 matched controls from six prospective cohorts. The proposed study will incorporate a rigorous two-stage design to screen and validate circulating metabolite biomarkers for gastric cancer risk in East Asians and European Whites (Aim 1 & 2). Furthermore, we will assess mediating effects of the identified metabolites on the associations between known risk factors and gastric cancer to facilitate the understanding of involved biological mechanisms (Aim 3). Through aseries of rigorously designedresearch activities, fresh insights into mechanisms of gastric carcinogenesis and novel biomarkers for future development of risk assessment tools are expected.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY – OVERALL In the U.S. and other high-income countries, immunotherapy is transforming the management of some cancers, including colorectal cancer (CRC). However, these advances remain largely inaccessible to patients in low- and middle-income countries (LMICs), despite evidence suggesting they may derive substantial benefit. Our long-standing collaborations in sub-Saharan Africa have highlighted significant differences in CRC outcomes – 53% of Nigerian CRC patients will die within a year of diagnosis, compared to 17% in the U.S. – and incidence in Nigeria is rising. In exploring the biological basis for these outcome differences, we discovered a threefold higher prevalence of microsatellite instability-high (MSI-H) CRC among Nigerian patients compared to U.S. counterparts. This suggests a distinctive tumor immunobiology and a compelling rationale for therapeutic exploration. Based on these findings, we are separately launching the first prospective trial of immunotherapy for MSI-H CRC in sub-Saharan Africa. While promising, this work has illuminated the broader lack of immuno-oncology data from African populations. This data gap limits not only local clinical application, but also the generalizability of immunotherapy strategies globally, including in African-descended populations in the U.S. Bringing the promise of immunotherapy to these populations requires a contextual understanding of CRC immunobiology, biomarkers to efficiently and cost-effectively select patients most likely to benefit from immunotherapy, and regional immuno-oncology experts with knowledge of immunotherapy and immune-related side effects. To address these needs, we will establish the Nigerian Immuno-Oncology Research (NOLA) program, leveraging our existing strengths in global cancer research and immuno-oncology to create a focused, coordinated initiative. Specifically, in this P20 project, we will unite a multi-disciplinary group to 1. generate pilot data to guide future studies investigating the tumor microenvironment of Nigerian CRC; 2. build a collaborative, multi-institutional research platform that integrates immuno-oncology efforts across Nigerian and U.S. cohorts; and 3. establish CRC biobanks and databases to support future immuno-oncology research. Training opportunities embedded in this project foster cross-cultural, systems-level thinking among U.S.-based early-career investigators, thereby expanding the breadth and resilience of the U.S. biomedical workforce. This program strengthens the U.S. cancer research enterprise by developing collaborative platforms that enhance biological insight, global preparedness, and scientific generalizability, particularly in populations of African ancestry who are insufficiently represented in existing datasets but significantly affected by cancer morbidity in the U.S.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY Chimeric antigen receptor (CAR) T-cell therapy has transformed the treatment of hematologic malignancies, resulting in remarkable disease response rates. Yet, the unprecedented efficacy of CAR-T cell therapy is limited by unique toxicities. Among them, immune effector cell-associated neurotoxicity syndrome (ICANS), a specific type of neurotoxicity, found in up to 70% of patients treated with CAR-T cells, consists of neurologic symptoms including encephalopathy, aphasia, seizures, which often develop acutely. While most cases of ICANS appear to resolve within the first month post-CAR-T therapy, a subset of patients develop delayed or chronic neurologic symptoms. However, the acute and chronic neurologic sequelae of CAR-T therapy are poorly understood, making this an urgent and unmet clinical need. Our preliminary data suggest that the KYN pathway is a critical link between systemic inflammation and acute ICANS. Also, changes in blood markers of neuroaxonal and glial injury have been reported during ICANS, suggesting that these are relevant markers. Our ongoing pilot study assessing neurocognitive and quantitative neuroimaging outcomes in lymphoma patients treated with CAR-T cell therapy suggest a decline in verbal memory as well as decreases in white matter and resting state functional connectivity from pre- to 3-4 months post-CAR-T. Taken together, these findings suggest that a deeper understanding of ICANS and its chronic impact is of utmost relevance given the growing use of CAR-T therapy. Aim 1 will investigate the molecular pathogenesis of acute ICANS using our large biobank of longitudinally collected samples from 135 lymphoma patients treated with CAR-T cells. We will assess changes in the KYN pathway neuroactive metabolites, immune cell transcriptomes, proinflammatory proteins, and markers of neuronal and glial damage, pre- and within the first month post-CAR-T, to explore their association with the development of acute ICANS. Aim 2 will assess neurocognitive function and structural and functional neuroimaging prospectively in 120 additional lymphoma patients, pre-treatment, as well as 3- and 6-months post-CAR-T, and their association with acute ICANS. We will explore acute and longitudinal changes in metabolites, proinflammatory cytokines, and neuronal/glial damage markers, and their association with neurocognitive and neuroimaging outcomes. This project involves a multi-disciplinary team with complementary expertise in neurology, immunology, neuropsychology, neuroradiology, neuroimaging, bioinformatics, and biostatistics with extensive experience in cancer research. We are uniquely positioned to conduct this ambitious research plan, and our results will offer novel insights into the pathogenesis of acute neurotoxicity and provide phenotypic characterization of chronic neurologic sequelae after CAR-T cell therapy. By improving the current scientific understanding and clinical management of ICANS, our study will contribute to identifying patients at risk for chronic neurotoxicity, improving clinical decision-making, and developing interventions to prevent/minimize treatment-related sequelae without decreasing CAR-T therapy efficacy.

NIH Research Projects · FY 2025 · 2024-09

ABSTRACT/SUMMARY Two rare types of thyroid cancer currently have extremely poor outcomes. Radioiodine-refractory (RAIR), recurrent and/or metastatic (R/M) differentiated thyroid cancer (DTC) is incurable and has a 10-year survival of ~10%. Anaplastic thyroid cancer (ATC) is a highly aggressive, dedifferentiated thyroid cancer with a median survival of <6 months from the time of diagnosis. Both tumors are driven by mutually exclusive genetic alterations in the mitogen-activated protein kinase (MAPK) pathway. Targeted therapies against BRAFV600E and receptor tyrosine kinase (RTK) rearrangements (RET, NTRK, or ALK) have revolutionized outcomes for these genomic subsets of RAIR DTC and ATC. However, developing therapies for RAS, NF1, and non-V600 BRAF mutations, which activate MAPK signaling through RAF dimers (“dimer-driven”), has been significantly more challenging. Avutometinib is a next-generation MEK 1/2 inhibitor that inhibits both MEK and RAF to overcome the rebound pathway stimulation that limits the activity of other MEK inhibitors against dimer-driven disease. This has translated to promising clinical activity with avutometinib against RAS-mutant tumors. Nonetheless, on-treatment biopsies from patients have revealed that avutometinib treatment increases focal adhesion kinase (FAK) phosphorylation/activation, which has been well established to mediate intrinsic resistance to MAPK pathway inhibition in several tumor types. Subsequent randomized data has confirmed avutometinib plus the FAK inhibitor defactinib produces superior clinical efficacy compared to avutometinib alone in patients with low-grade serous ovarian carcinoma, including RAS-mutant tumors. These data, along with evidence from our studies of patient tumors and genetically engineered mouse models of advanced thyroid cancer, serve as the clinical and biologic basis for our phase II trial of avutometinib plus defactinib in dimer-driven thyroid cancers. To expeditiously identify promising clinical signals, we have developed an innovative basket trial design which will allow us to evaluate in both dimer-driven RAIR DTCs and ATCs the contribution of 3 putative, non-exclusive mechanisms to avutometinib + defactinib efficacy: tumor cell-autonomous antitumor effects, redifferentiation of tumors to restore and/or enhance RAI sensitivity, and modulation of the immune tumor microenvironment (TME). In Aim 1, we will evaluate overall response (primary) as well as progression-free survival and redifferentiation rates/efficacy (secondary). In Aim 2, we will interrogate in pre- and on-treatment biopsies the impact of avutometinib + defactinib upon MAPK- and FAK- regulated signaling and analyze how it relates to clinical response and redifferentiation. In Aim 3, we will analyze how combined MAPK and FAK pathway inhibition modulates the immune TME, specifically testing the hypothesis that inhibiting these signals will deplete immune-repressive tumor-associated macrophages (TAMs) in dimer-driven ATCs. This trial will identify the optimal application(s) for this novel combination in patients with advanced thyroid cancers to inform future drug development efforts.

NIH Research Projects · FY 2024 · 2024-09

Project Summary Despite advances in treatment, sarcomas like rhabdoid tumors and desmoplastic small round cell tumors (DSRCT) remain challenging, with frequent therapy resistance and fatal outcomes. The rarity of these cancers (<1 case per 1,000,000 people per year for each type) and the limitations of current genomics and proteomics methods underscore a critical need for innovative approaches to discover improved targets for therapy. Our project aims to propel the study of ultra-rare cancers by integrating cutting-edge genomic and proteomic technologies to uncover tumor-specific biological mechanisms and therapeutic targets, as envisioned by the FDA Oncology Center of Excellence. Our overarching goal is to elucidate the fundamental biological processes and molecular mechanisms driving cancer pathogenesis, specifically for rare cancers such as rhabdoid and DSRCT tumors. By defining their tumor-specific proteomes, we seek to enable the discovery of specific therapeutic targets to improve the survival rates of children and adults facing these refractory cancers. Leveraging a unique MSK cohort of over 80,000 diverse patients with active disease, including those with rhabdoid and DSRCT tumors, we propose a synthetic integration of long-read sequencing of bulk DNA, bulk RNA and single cell RNA, and high-resolution multidimensional mass spectrometry proteomics. This approach will enable us to construct comprehensive maps of tumor-specific cell surface proteomes, revealing neomorphic gene products and non-canonical protein isoforms. In Aim 1, we will delineate tumor-specific neomorphic and non-canonical cell-surface gene transcripts through the combined use of integrative long-read and single-cell sequencing. Aim 2 will define the expression of neomorphic cell surface proteins directly using integrative mass spectrometry proteogenomic approaches, focusing on non-canonical proteoforms and their tumor-specific post-translational modifications. This approach includes rigorous technical validation measures and aims to contribute to the NIH Cancer Data Science Initiative’s vision for an open data cancer ecosystem. By addressing the critical gap in current cancer research methodologies and focusing on rare rhabdoid and DSRCT tumors, this project stands to make a significant impact on the understanding and treatment of ultra-rare cancers, paving the way for the development of targeted therapies and improving patient outcomes.

NIH Research Projects · FY 2025 · 2024-09

SUMMARY Our overarching goal is to develop a novel theranostic prostate-cancer-specific CAR-T cell paradigm that will allow for successful targeting of antigen-heterogeneous neuroendocrine prostate cancer (NEPC) under image guidance. Our proposal and experiments are based on the observation that NEPC develops in most cases through increasing de-differentiation from the clinical state of castrate-resistant disease; along this trajectory, DLL3 expression increases and PSMA expression decreases on tumor cells. Hence, targeting DLL3 as well as PSMA with a bi-specific immunotherapeutic is a logical choice when developing a novel targeted therapeutic approach for this disease state. Our approach takes advantage of existing institutional expertise in developing CAR-T-cell-based therapeutic strategies following very encouraging clinical outcomes in hematologic malignancies and in mesothelioma. Here, we aim to extend this approach to metastatic NEPC. We realize that translation of the clinical CAR-T cell success in hematologic malignancies to metastatic solid tumors must overcome several obstacles. It necessitates systemic delivery of CAR-T cells, and any therapeutic success will require the delivery of CAR-T cells to metastatic sites in sufficient quantity, their survival and potent effector function at these sites, and the prevention of tumor immune escape due to plasticity and heterogeneity of targeted surface antigens. Therefore, we have designed a series of experiments to address these challenges in logical sequence. Our research team combines expertise in CAR-T cell design, development of T-cell tracking techniques, translation of novel imaging probes in prostate cancer, and clinical trial design. We will conduct a series of pre-clinical studies in order to refine the technology and prepare for translation in patients with NEPC. We will begin by evaluating the therapeutic efficacy of DLL3-specific CAR-T cells in NEPC models and determining the lower DLL3-expression threshold on cancer cells for CAR-T tumoricidal effect. Next, we will develop DLL3/PSMA-bi-specific CAR-T cells and evaluate their ability to eradicate tumors with heterogeneous DLL3 and PSMA expression using real-time multi-reporter bioluminescent imaging. To facilitate clinical translation of this approach, we will verify that these bi-specific CAR-T cells can be tracked by PET imaging with the clinically approved 18F-MFBG radiotracer, and their efficacy predicted and measured using the clinically established DLL3- and PSMA-specific radiotracers 89Zr-DFO-SC16 and 68Ga-PSMA, respectively. In summary, this proposal brings together translatable CAR-based therapeutic and imaging components in genetically engineered T-cells, allowing for non-invasive monitoring of CAR-T cell trafficking, tumor targeting, and tumor response. Of note, the theranostic paradigms that will be developed and validated in this proposal can also be translated and used to address comparable clinical questions in other DLL3-positive malignancies.

NIH Research Projects · FY 2025 · 2024-09

Precision oncology is a firmly established pillar in the practice of cancer medicine, but we now recognize new challenges to its broad implementation. These include: (i) heterogeneity in response to precision oncology drugs in patients with identical driver mutations, (ii) differences in driver mutation frequencies in patients from different ethnicities, with implications for ensuring optimal treatment, (iii) increased subtyping of cancer into hundreds of “rare” diseases, and the resultant operational challenges in clinical trial design and execution, and (iv) a limited understanding of how to effectively leverage observational (real world) data to address these challenges. The investigators in this P01 Program have had a longstanding collaboration over the past 8 years, helping to build a large (>148,000 patients), international clinico-genomic cancer registry known as AACR Project GENIE. We have come together to investigate these issues and propose four highly integrated projects related to these themes. Project 1 seeks to overcome methodologic barriers in the analysis of observational clinico-genomic data. Project 2 will address the role of genetic ancestry in precision oncology outcomes and potential inequities in how precision oncology diagnostic tests are developed. Project 3 will use real world evidence to inform clinical decisions for the treatment of cancer patients that cannot be addressed using conventional clinical trial datasets and will optimize the reporting of these findings using the OncoKB knowledge base. The projects will be supported by the Curation and Statistical Analysis Core (data abstraction and biostatistics support); the Molecular Pathology and Bioinformatics Core (molecular profiling and data capture); and the Administrative Core (incorporating existing GENIE infrastructure for data sharing, communications, and administrative support). Our proposal is highly synergistic as it brings together a multi-institutional team of distinguished investigators in population science, population genetics, cancer genomics and experimental therapeutics, with a substantial track record of collaborative interactions, who will work together to address these important topics in precision oncology.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY Over 500,000 survivors of childhood cancer are living in the United States today. Cardiovascular disease is the leading cause of noncancer late mortality in this cohort. Subgroups of survivors, including individuals previously exposed to radiation to the chest, abdomen, and total body, are also at treatment-related risk for a constellation of cardiometabolic comorbidities, including type 2 diabetes mellitus, chronic inflammation, and hypertension, which further amplifies their risk of cardiovascular disease. Tailored, scalable and sustainable interventions that target modifiable risk factors, such as obesity, are thus urgently needed to reduce cardiometabolic risk in this cohort. Time-restricted eating, or voluntary abstinence from all caloric intake for 14- 16 hours per day, has been demonstrated in pre-clinical and human studies to combat these risks. This mode of fasting appears to be feasible, safe, and sustainable in diverse populations. Results on the efficacy of time- restricted eating, however, are discrepant across studies, with some studies demonstrating marked improvement in weight, glycemic control, and circadian rhythm alignment and others showing minimal effect. Yet this intervention has not been tested in childhood cancer survivors, and it is unknown whether it would be an efficacious risk-reducing intervention in this population. The proposed STRENGTH (Survivors engaged in Time-Restricted EatiNG after THerapy) Study would fill this gap by determining the efficacy of time-restricted eating to reduce weight and improve markers of cardiometabolic risk in radiation-exposed survivors with overweight/obesity. To this end, we will conduct a phase IIb randomized controlled trial of 300 radiation- exposed survivors of childhood cancer with overweight/obesity and enrolled in the Childhood Cancer Survivor Study (CCSS). Participants will be randomized to either a 6-month time-restricted eating intervention followed by a 6-month maintenance phase, or usual care, which consists of mailed educational handouts on healthy lifestyle behaviors. We will: (1) Determine the impact of time-restricted eating on measures of weight status and other weight-related outcomes (waist circumference, percent body fat); (2) Assess the impact of time- restricted eating on cardiometabolic risk profile (blood pressure, glycemic control, insulin resistance, adipokines, markers of inflammation); (3) Identify moderating and mediating factors, as well as barriers and facilitators, associated with adherence to the intervention and weight loss, with the goal of informing future implementation efforts. The proposed time-restricted eating intervention, which is grounded in social cognitive theory, will include stepwise progression to build self-efficacy using “nudges” and motivational interviewing. This proposal will leverage novel mHealth technology to deliver a fully remote, theoretically grounded behavioral intervention to reduce cardiometabolic risk and thereby address the burdens of cancer and its treatment among adult-aged survivors of childhood cancer. Time-restricted eating may represent a sustainable intervention to improve long-term outcomes in a highly morbid survivor cohort.

NIH Research Projects · FY 2024 · 2024-09

7. Project Summary/Abstract Funds are being requested to purchase a Super Argus PET/CT 6R small-animal PET (positron emission tomography) and CT (transmission computed tomography) scanner (Scintica, Webster, TX) to replace our Inveon microPET/microCT scanner. Importantly, Siemens Medical Solutions (the manufacturer of the Inveon) has informed us that the support of the Inveon will end within the next 1 to 2 years (i.e., by the end of 2025), meaning that parts and service for needed repairs, software upgrades, and applications support will very likely not be available beyond that end-of-support date. The proposed Super Argus PET/CT 6R will be used in mouse and rat models to non-invasively and longitudinally: monitor cancer and other disease processes; assess trafficking of cancer, immune, and stem cells; evaluate cellular responses to investigational anti-cancer therapies and other interventions; assay genetic pathway expression; and measure tumor targeting and biodistribution of new radiodiagnostic and radiotherapeutic agents. Like the Inveon, the state-of-the-art Super Argus PET/CT 6R, to be installed in our existing Animal Imaging Core Facility, will be a critical, broadly used resource for the pre-clinical and translational oncology research program at Memorial Sloan Kettering Cancer Center (MSK). The Super Argus PET/CT 6R, as noted, will replace the ageing Inveon, installed in 2013 and now over 10 years. With its impending end of support, should the Inveon require service support and replacement parts, they likely will not be available and this heavily used, mission-critical resource would be immediately and permanently disabled - a devastating blow to our entire pre-clinical research program. Funds for a new, fully supported, higher-performance, and more functional small-animal PET/CT system, the Super Argus PET/CT 6R, are therefore urgently requested to maintain, enhance, and expand our PET and CT imaging capabilities. In addition to maintaining manufacturer service and applications support, among the critical advantages of the Super Argus PET/CT 6R over our existing, now-outdated Inveon are: (1) superior spatial resolution – 0.7 mm full-width half-maximum (FWHM) for PET and as low as 15 µm for CT; (2) phoswich detector-based depth-of-interaction (DOI) correction for PET and resulting constant resolution across the 120-mm transaxial field of view (FOV); (3) high PET sensitivity - 12%; and (4) a long axial FOV - 15 cm. The Super Argus PET/CT 6R is a fully integrated multi-modality system providing transparent registration of PET and CT studies or serving as stand-alone PET and CT systems. In addition to numerous new and ongoing funded projects which will use the Super Argus PET/CT 6R, our Facility and MSK have well-established technical and scientific expertise to continue to productively utilize this instrument and an efficient administrative and solvent financial infrastructure in place to support and maintain this instrument and its effective use long-term.

NIH Research Projects · FY 2024 · 2024-09

PROJECT SUMMARY/ABSTRACT The proposed study is significant because we aim to understand the cellular mechanisms that regulate lymph node (LN) maintenance in aging. Aging causes LN atrophy and fibrosis and is associated with impaired immune responses. However, although the morphological changes that occur in aged LNs have been well studied, the molecular mechanisms that cause these changes are poorly understood. This gap in knowledge is important because LNs play a key role in orchestrating immune responses. Therefore, our studies will help uncover therapies that may improve immune function in elderly men and women. Lymphatic endothelial cells (LECs) are the building blocks of the lymphatic system and, as a network of lymphatic vessels, transport of interstitial fluid and antigens to regional LNs. LECs are also a component of the LN stroma and regulate LN organogenesis during development and LN hypertrophy in response to inflammatory conditions. These regulatory pathways are coordinated by canonical and non-canonical nuclear factor kappa- light-chain-enhancer of activated B cells (NFkB) signaling in LECs. Aging is known to impair lymphatic function and transport of interstitial fluid/antigens to the LN, and we have shown in preliminary studies that LN LEC heterogeneity is altered in aged LNs and the expression of NFkB by LN stromal LEC is decreased in aging animals. We also have found that deleting canonical NFkB-Rel-A signaling in LECRel-A-/- mice results in significant LN atrophy and cellularity that is a phenocopy of the aging LN. Based on this rationale, our central hypothesis is that decreased lymphatic function in aging results in diminished expression of NFkB by LN stromal LECs, impaired production of lymphocyte survival signals, and aging-related LN atrophy. Our study is innovative because we plan to conduct single-cell RNA sequencing analysis of LN stromal LECs, and we have developed inducible transgenic Cre-lox mice in which we can selectively decrease LEC NFκB signaling in adult animals. This model system enables us to directly study the role of LEC NFκB expression in LN maintenance and atrophy. These transgenic mice, combined with our LN transplantation, and in vitro 3D models, will also enable us to selectively analyze changes in LN LEC NFκB signaling and understand how these populations contribute to adult LN homeostasis using two Specific Aims: (1) Determine the effects of aging on LN LECs and effect of lymphatic flow on LEC NFkB expression and (2) determine how LEC NFκB signaling regulates LN homeostasis in adult mice. At the conclusion of the proposed study, we expect to understand the cellular mechanisms that contribute to aging-related LN atrophy and adult LN homeostasis. This understanding will provide the basis for future studies designed to prevent or reverse aging-related LN atrophy.

NIH Research Projects · FY 2025 · 2024-09

The desmoplastic small round cell tumor (DSRCT) remains one of the most lethal adolescent/young adult (AYA) sarcomas. The EWSR1-WT1 gene fusion is the primary and defining genetic driver alteration in this sarcoma but a better understanding of EWSR1-WT1 oncogenic mechanisms is needed to identify novel, rational therapeutic strategies that will be more effective. We propose to address these knowledge gaps by establishing the first genetically engineered mouse model of DSRCT and to define its pathobiology more fully by harnessing two recent novel observations in DSRCT, namely induction of neotranscripts and recurrent ARID1A loss as a frequent additional genetic alteration in these cancers. Aim 1: Developing a mouse model of DSCRT by germline and somatic genome editing. We will use the in vivo somatic chromosomal engineering method pioneered by our laboratory (A.V.) to induce the EWSR1-WT1 translocation in mice, based on co-expression of Cas9 and two gRNAs targeting the desired translocation breakpoints. As p53 is mutated in approximately 10% of DSRCT cases, we will induce the translocation concomitantly with p53 deletion in a cohort of mice to enhance tumor development. Prompted by the recent finding that ARID1A mutations are commonly observed in DSRCT patients (see Aim 3), we will also test the consequences of Arid1a loss on DSRCT initiation and progression in vivo. Aim 2. Defining the landscape of EWSR1-WT1-associated neotranscripts in DSRCT. We (J.W., O.D.) have recently reported that, in addition to dysregulating the expression of many target genes, oncogenic fusions such as EWSR1-WT1 also drive the expression of completely novel sequences which we have termed neotranscripts (J.W., O.D.). These are unannotated multi-exonic, polyadenylated RNAs not expressed in any healthy tissue but directly induced by the altered localization patterns and co-factor associations of the oncogenic fusion protein. A subset of neotranscripts are actively translated into protein products, representing potential neoantigens. We will build upon our initial discovery of these sequences and use them as a model system for dissecting EWSR1-WT1 mediated gene regulation as well as determining what role these neotranscripts or their encoded proteins play in tumor biology. Aim 3. Determining the role of recurrent ARID1A loss in DSRCT. While DSRCT displays one of the lowest somatic mutation rates of any human solid cancer, intriguingly, inactivation of ARID1A, a subunit of the canonical BAF complex, is the most common secondary mutation in DSRCT (about 6-12% of cases), being more common than even TP53 alterations (M.L.). We propose to use isogenic models of ARID1A loss in DSRCT cell lines to define the phenotypic and epigenetic effects of this recurrent secondary alteration in DSCRT and use chemical and functional genomic screens to identify novel vulnerabilities associated with ARID1A loss which may also yield more general insights into DSRCT pathobiology.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY/ABSTRACT Pelvic radiotherapy is the standard of care for numerous gynecological and anorectal cancers affecting women. Post-operative intravaginal brachytherapy (IVRT) is used to prevent cancer recurrence at the vaginal cuff for high-intermediate risk endometrial cancer patients. IVRT delivers conformal radiation dose to the vagina cuff and canal. This results in high rates of acute and late vaginal toxicities, such as vaginal stenosis, which is defined as the abnormal tightening and shortening of the vagina due to fibrosis. These effects can lead to debilitating pain, sexual dysfunction, and poor quality of life. Current clinical assessments of vaginal toxicity are subjective and based on clinician and patient reported grading scales that leads to variability. There is a clinical need for objective measurements of vaginal tissue health, and we propose a multiparametric ultrasound imaging approach for vaginal tissue characterization before and after IVRT. In this study, we will develop a system and methodology for three-dimensional (3D) multiparametric imaging of the vaginal wall using B-mode, ultrasensitive microvessel imaging (UMI), shear-wave elastography (SWE), and tissue microstructure characterization with quantitative ultrasound (QUS). As a secondary objective, we will test the system in a clinical study with fifteen endometrial cancer patients receiving IVRT to identify associations between multiparametric ultrasound metrics and current clinical assessments of vaginal toxicity. This will be the first study using a multimodal ultrasound approach for characterization of vaginal tissue. We hypothesize that multiparametric ultrasound will be produce metrics that are sensitive to anatomical changes (B-mode), microvascular damage (UMI), fibrosis stage (SWE), and tissue microstructure and cellular death (QUS). There is increasing awareness of radiation-induced toxicities and considerable effort has gone into strategies to reduce toxicities and improve overall quality of life and sexual functioning. Multiparametric ultrasound imaging offers a non-subjective, quantitative approach for detecting and characterizing vaginal health, which is typically overlooked with current medical imaging. Ultimately, this approach will lead to predictive imaging biomarkers that clinicians can use as outcome metrics of novel treatment strategies and interventions aimed at preventing and treating radiotherapy-induced vaginal toxicities. If successful, the research will introduce a safe, cost- effective imaging platform to improve the health of women suffering from this chronic gynecological condition.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY/ABSTRACT Hematopoiesis ensures the continuous formation and turnover of blood cells during homeostasis, which shifts drastically towards production of inflammatory innate immune cells upon infection. At steady state, myeloid progenitors express a default signaling pathway essential for monocyte/macrophage production and require extrinsic signaling and transcription factors to direct lineage commitment towards other myeloid cell fates. During infection, coordination of cytokines and growth factors shifts myelopoiesis dramatically towards neutrophilia, a process known as emergency granulopoiesis. For instance, neutrophil production during systemic infection is boosted by over 10-fold, increasing cell numbers to over 1012 in a single day. Therefore, the rapid response to lineage-specific cytokines and growth factors to drastically increase granulocyte output is essential for pathogen clearance and host survival. Inappropriate lineage fidelity is exemplified by aberrant responses to infection observed in systemic inflammation and septic shock, however, the cellular machinery necessary to shift lineage commitment to emergency neutropoiesis remain poorly understood. Recently, we made the surprising discovery that the chloride-sensing kinase with-no-lysine 1 (WNK1) is indispensable for myeloid progenitor fate, with myeloid-specific deletion of Wnk1 incurring dramatic loss of tissue-resident macrophages, disrupted organ development, systemic neutrophilia, and mortality by 4 weeks old. Mechanistically, the cognate CSF1R cytokine, macrophage-colony stimulating factor (M-CSF), triggers macropinocytosis in myeloid progenitors, which in turn induces phosphorylation and activation of WNK1. Absence of WNK1 or macropinocytosis inhibition drove myeloid progenitor differentiation into granulocytes in vitro and in vivo. The goal of this proposal is to mechanistically dissect the WNK1-dependent pathway initiated by M-CSF-induced macropinocytosis and to investigate how this pathway is systematically shifted during emergency granulopoiesis. During the mentored phase of this application, I will master new techniques such as subcellular fractionation and macropinosome isolation. I will study how to analyze, integrate, and communicate complex functional genomics and metabolomics data. Together with colleagues, we will develop new tools to assess the importance of macropinocytosis during hematopoiesis in vivo, allowing us to understand how the contents of macropinosomes inform myeloid lineage commitment. With the guidance of my mentoring committee, I will strengthen my scientific and professional skillsets in preparation for the independent phase. During the independent phase, I will explore targeting the WNK1 pathway during sepsis using small molecules. Further, I will combine my expertise in in vivo sepsis models, emergency granulopoiesis, and neutrophil biology together with tools developed during the mentored phase to understand how macropinocytosis influences the innate immune response. The proposed project will allow me to successfully establish an independent academic research program.

NIH Research Projects · FY 2025 · 2024-09

Recent studies have discovered that where one lives is associated shorter or longer breast cancer (BCa) recurrence-free survival (RFS) independent of individual, behavior, lifestyle, tumor, and National Comprehensive Cancer Network guideline-concordant treatment characteristics, suggesting unaccounted mechanisms by which one’s geographic location impacts BCa RFS. Our preliminary data indicate that geographic location, through neural influences on BCa biology, may lead to shorter or longer BCa RFS. Specifically, we leveraged the novel genomic-epidemiologic infrastructure of our Miami Breast Cancer Cohort study, a prospective cohort study with clinically and survey-annotated blood and tissue samples, to discover that positive and negative factors associated with where one lives and positive and negative perceptions of where one lives are independent predictors of 1) lower or higher levels of myeloid lineage immune cell activation in blood (a pattern known as the Conserved Transcriptional Response to Adversity; CTRA) and 2) a less or more aggressive BCa molecular biology profile (up-regulation of proinflammatory transcription factors (TF) NF-kB and AP-1), respectively, after controlling for individual, tumor, and guideline-concordant treatment factors. In addition, an aggressive BCa molecular biology profile correlated with shorter RFS. We also discovered that where one lives and one’s perceptions of where he or she lives, are associated with up- or down-regulation of TF CREB [a marker of sympathetic nervous system (SNS) activity], implicating the SNS as a potential mediator of geographic location and RFS through NF-kB and AP-1 regulation. Combined, our preliminary data provide strong evidence for our central hypothesis that where one lives and one’s perceptions of where he or she lives are associated with RFS, in part, through biologic mechanisms mediated by SNS regulation (CREB) of circulating myeloid lineage immune cells (CTRA) resulting in a more or less aggressive tumor biology molecular profile. To test our central hypothesis, we propose the following aims in the largest social genomics BCa study to date, consisting of 750 BCa patients enrolled in our study: Aim 1) Define the contributions of where one lives (geographic location) and one’s perception of where he or she lives on BCa RFS. Aim 2) Validate the biologic mechanisms mediating the effects of geographic location on RFS. Aim 3) Evaluate the spatial effects of clinical and biologic correlates of shorter RFS. Combined, these findings, to our knowledge, will represent the single largest social genomics and geospatial BCa research effort to how geographic location influences BCa RFS with an innovative focus on 1) isolating the distinct biologic mechanisms associated with where one lives, one’s perceptions of where he or she lives, and RFS, and 2) geospatial identification of areas with less aggressive biology and longer RFS to inform future policy and targeted behavioral or therapeutic interventions to reduce SNS activity and thus based on prior randomized control trials, reverse aggressive tumor biology, ultimately improving BCa RFS.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY/ ABSTRACT Immune checkpoint blockade (ICB), aimed at reinvigorating immune cells such as T-cells, has exhibited clinical success in a subset of patients yet fails to be applicable to many tumor subtypes. These unresponsive subtypes upregulate gene signatures associated with amino acid (AA) metabolism and deprivation. However, how AA deprivation contribute to poor ICB response remains unclear. Upon chronic antigen stimulation, T-cells become “exhausted”, an alternative differentiation state that entails the loss of cytotoxic effector function and proliferative capacity. T-cell exhaustion involves extensive transcriptomic and epigenetic remodeling; however, the loss of cytokine production occurs despite adequate expression of transcripts encoding cytokines, suggesting post- transcriptional mechanisms of restricting effector function within tumors. The long-term goal is to elucidate the mechanisms underlying T-cell dysfunction within the tumor microenvironment and leverage these insights to enhance ICB efficacy. The predoctoral research (Aim 1) will aim to investigate how local AA availability limits effector function in tumor-infiltrating T-cells. Preliminary data showed that intratumoral T-cells can not engage in efficient translation and they experience glutamine deprivation in the tumor microenvironment. Elevated translational demand downstream of chronic T-cell receptor signaling cannot be met when extracellular AAs are limiting, restricting both global translation rate and cytotoxic cytokine production. Specific Aim 1.1 will seek to determine how local AA availability impacts the exhausted T-cell proteome. Nascent transcriptomic and translatomic alterations will be profiled to evaluate the impact of AA limitation on gene-specific translation rates, followed by ribosomal footprinting assays to identify cell state- and AA-dependent stalling in vivo. Specific Aim 1.2 will investigate the impact of enhancing AA availability on ICB. A broad-spectrum AA transporter will be overexpressed in T-cells to examine whether it enhances tumor control in response to ICB and overcomes the immunosuppressive effects conferred by cancer-associated fibroblasts via restricting intratumoral AA availability. My postdoctoral research (Aim 2) will focus on the role of non-coding RNAs in translational suppression during terminal T-cell exhaustion. I will profile how T-cell exhaustion impact the expression of non-coding RNAs and examine whether non-coding RNA subsequently modulates translation and cytokine production. Overall, these two projects will unveil the distinct mechanisms driving T-cell dysfunction through translational suppression during early tumor-infiltration and late terminal exhaustion. The research and training plan outlined in this proposal will be completed with the joint mentorship of Dr. Santosha Vardhana and Dr. Jayanta Chaudhuri at Memorial Sloan Kettering Cancer Center (MSK). MSK’s top-notch cancer research environment and abundant resources in conjunction with the support of the Gerstner Sloan Kettering Graduate School will guarantee the successful completion of the proposed research and career development plans.

NIH Research Projects · FY 2025 · 2024-08

PROJECT ABSTRACT We developed and translated CD28 costimulated chimeric antigen receptor (M28z CARs) T cells that target mesothelin (MSLN), a cancer-associated antigen that is overexpressed in a majority of solid tumors; 65 patients have been treated in phase I/II trials, with no on-target, off-tumor toxicity. Having demonstrated that regionally administered CAR T cells avoid pulmonary sequestration and are enhanced by early antigen-activated CD4 CAR T-cell helper function, we delivered M28z CAR T cells intrapleurally in patients with malignant pleural mesothelioma (MPM), promoting tumor infiltration. To overcome antigen stress–induced T-cell exhaustion, we treated 27 patients with MPM with an anti-PD1 agent after a single dose of intrapleural M28zCAR T cells (median survival 18 months vs. 10 months following second line pembrolizumab). To facilitate tumor-specific checkpoint blockade and to achieve functional persistence without repeated administration of anti-PD1 agent, we translated a CAR T-cell–intrinsic PD1 dominant negative receptor (PD1DNR); 8 patients have been treated, with no CAR- or PD1DNR-related toxicities. Recognizing the necessity of the IFNγ pathway for solid tumor killing by CAR T cells, we exploited a c-KIT mutation, D816V (KITv), as a costimulatory domain (Nat Cancer 2023). A single infusion of M28zKITv CAR T cells was safe, resisted TGFβ-mediated suppression, prolonged survival, and demonstrated functional persistence in multiple solid tumor models including in low-antigen expressing tumors; clinically available kinase inhibitors are effective as on/off, tunable safety switch. Herein, we hypothesize that, whereas IFNγ-mediated cytotoxicity is the predominant mechanism of KITv CAR T cells against high-antigen-expressing targets, for low- or non-antigen- expressing targets, IFNγ-mediated upregulation of ICAM1 on tumor cells facilitating LFA1(CAR T-cell)-ICAM1 synapse (increased avidity) and apoptosis of IFNγ receptor–expressing target cells are the respective biologic mechanisms (Aim 1), KITv-induced IFNγ potentiates the cytotoxicity of CD4 41BBz KITv CAR T cells against low-antigen-expressing targets and that CD28 costimulation promotes the helper function of CD4 28zKITv CAR T cells; however, PD1DNR is essential for functional persistence (Aim 2), and that PD1DNR-mediated checkpoint blockade extends beyond PDL1-expressing tumor cells and counteracts PDL1-overexpressing M2 macrophages (Aim 3). The significance of our approach lies in its effective combination of a solid tumor–specific scFv that is on target and safe (MSLN), costimulatory domains (KITv, CD28), checkpoint blockade (PD1DNR), and tumor infiltration (regional delivery). The impact of our proposal extends beyond pleural cancers; the majority of aggressive solid tumors express MSLN. The proposal’s innovations include— exploiting IFNγ-mediated cytotoxic mechanisms in a heterogenous antigen-expressing solid tumor, increased CAR T-cell avidity through a second synapse, potentiating CD4 CAR T-cell helper function to augment CD8 CAR T cells and counteracting PDL1-expressing M2 macrophages by the CAR T-cell intrinsic PD1DNR.

NIH Research Projects · FY 2025 · 2024-08

PROJECT SUMMARY/ABSTRACT Research. Non-small cell lung cancer (NSCLC) is the world’s deadliest cancer, but patients with NSCLC can have dramatically different outcomes, illuminating an urgent clinical unmet need for improved risk stratification. Our study is motivated by the following unresolved questions in NSCLC oncology: 1) What is the likelihood of recurrence for patients with definitively treated disease? 2) Which patients with advanced disease are most likely to benefit from consolidative radiotherapy? 3) What is the likelihood that a patient will develop central nervous system metastasis? We contend that predictive models derived from real-world data collected as part of standard of care, including tumor genomic profiling, imaging, and clinician notes, combined with newer clinical assays such as circulating tumor (ct)DNA sequencing and radiomics will advance personalized answers to these questions, leading to improved outcomes for patients. We have recently developed methods to overcome barriers to using real-world data with transformer-based natural language processing, eliminating the need for time-intensive manual curation of clinician notes, yielding structured data critical for developing predictive models. In a proof of principle study, we validated the prognostic value of ctDNA sequencing merged with radiomic, tumor registry and tissue genomic data to create a richly annotated dataset an order of magnitude larger than recent manually curated cohorts. Our preliminary studies show that multimodal models incorporating complementary data streams improve overall survival prediction over any single data modality, such as stage or tissue genomics, and standard of care biomarkers. Based on these results, we hypothesize that specific combination models, encompassing real world data from ctDNA and clinicogenomic sources, more accurately inform tumor biology and patient outcomes than single-modality variables. We will improve risk stratification and clinical management of NSCLC by studying whether and how real-world data can be used to develop multimodal risk models that in the future could be deployed in clinical settings with minimal patient and clinician overhead. Candidate. Justin Jee, MD PhD is an Instructor in the Thoracic Oncology Service at MSK. His goal is to integrate AI-extracted clinicogenomic data to discover multimodal biomarkers of antineoplastic response for patients with cancer. He will undergo a five-year training period with a multidisciplinary mentorship team including experts in computational oncology, machine learning, genomics, natural language processing, radiomics, and thoracic oncology to obtain the skills necessary to become an independent, tenure-track physician scientist. Environment. MSK is an academic cancer center renowned for patient care, innovative research, and training for junior faculty seeking careers as independent physician-scientists. MSK is home to MSK-IMPACT, an FDA- authorized, tumor/normal sequencing assay with over 100,000 samples sequenced to date, and MSK-ACCESS, a 129-gene liquid biopsy assay. Both assays are leveraged extensively in this proposal.

NIH Research Projects · FY 2025 · 2024-08

Summary DNA replication stress is a major source for genome instability associated with numerous diseases, including cancer. Therefore, understanding the molecular basis of replication stress and, conversely, how accurate, complete, and rapid chromosomal DNA replication is achieved during normal cell proliferation, is crucial for understanding the mechanisms that maintain or threaten genome stability during normal development and disease, respectively. Here we propose experiments that will illuminate both the intrinsic mechanism of the eukaryotic DNA replication machinery and its response to diverse replication stress conditions. Previously, we have generated a fully reconstituted origin-dependent DNA replication system based on purified budding yeast proteins. More recently, we have begun to reconstitute replisomes with purified human proteins. These systems form the central platform for research in our lab. The normal and uninterrupted progression of replication forks is frequently challenged by physical obstacles on the parental DNA. These include non-B-form DNA secondary structures and tightly bound non- histone protein-DNA complexes. While non-B-form DNA secondary structures induce fork stalling in a stochastic and unscheduled manner, the programmed stalling of the replisome at tightly bound protein-DNA complexes can serve important biological functions. How individual physical obstacles on chromosomal DNA induce fork stalling, are overcome by fork restoration mechanisms, or elicit distinct biological outcomes is incompletely understood. To illuminate fork stalling mechanisms, we will investigate the molecular mechanisms by which such physical obstacles impede the eukaryotic replicative DNA helicase, CMG (Cdc45- MCM-GINS). Specifically, we will determine the mechanisms by which G-quadruplexes (G4s) impede CMG progression on the leading strand and the mechanisms by which replication forks stalled at G4s may be restored by accessory DNA helicases and fork remodelers. As an example for programmed fork stalling at protein-DNA complexes, we will investigate the mechanism of unidirectional fork stalling at the replication fork barrier (RFB) derived from both yeast and human rDNA repeats and examine the mechanism of replication termination upon fork convergence at these sites. Stalled replication forks are both inducers and targets of the replication checkpoint, which is essential for the maintenance of genome integrity by stabilizing stalled forks and coordinating fork restoration with cell cycle progression. Continuing our previous work, which has led to the identification of yeast replisomes as direct targets for the checkpoint effector kinase, Rad53, we will determine the mechanism(s) by which the checkpoint controls fork progression.

NIH Research Projects · FY 2026 · 2024-08

Twenty-six million people in the United States have limited English proficiency, meaning that they report speaking English less than “very well”; this population is more likely to have poor cancer outcomes, in part as a result of communication problems. Medical interpretating services are mandated by laws and regulations, but their availability is limited by expense and logistical factors, which may potentially be overcome with technology-enabled solutions. We will compare three types of technology-enabled human-delivered interpreting solutions in a study with breast cancer patients with limited English proficiency who speak Spanish or Mandarin (1. “UN-style” simultaneous interpretation versus 2. audio consecutive, and 3. video consecutive interpretation) to determine which type has the lowest error rate, has the best patient outcomes, and is the most efficient, and what factors affect the implementation of these solutions; we will also explore the potential of artificial intelligence to deliver automated “UN-style” simultaneous medical interpreting, which could be a scalable and inexpensive future solution to providing medical interpreting services to people with cancer.

NIH Research Projects · FY 2025 · 2024-08

PROJECT SUMMARY A detailed understanding of B cell and antibody responses to placental versus foreign antigen will illuminate how dysregulated responses contribute to pregnancy complications and inform the design of vaccines to best protect mothers and children from infection. Our published work and new preliminary data highlight two checkpoints that restrain harmful B cell and antibody responses to the placenta. Checkpoint 1 involves how antigen interacts with B cells. Placental antigen is modified with sialic acid sugar-carrying glycans (sialoglycans) and activation of placental-specific B cells is suppressed in a mechanism that is at least partially dependent upon recognition of sialoglycans by inhibitory receptors on the B cell surface. Checkpoint 2 is a fail- safe mechanism that protects the conceptus from antibody-mediated attack. This “effector phase” checkpoint is illustrated by our observation that pregnancy remains unaffected even when high titers of anti-placental antibodies, experimentally introduced before mating, are present. Notably, both checkpoints are specific to placental antigen itself. Checkpoint 1 appears to be driven by immunosuppressive sialoglycans that coat placental antigen. For Checkpoint 2, our data and the literature support the hypothesis that anti-placental antibodies become modified with suppressive glycans that diminish their function. Because the checkpoints are placental antigen-specific, maternal B cell responses to foreign antigen remain intact. Our data show that to foreign antigen, pregnant mice generate B cell and antibody responses commensurate with non-pregnant mice. However, to our surprise, humoral responses during pregnancy occur seemingly independent of robust follicular helper T cells (Tfh). Since humoral immunity in non-pregnant hosts is well documented to be regulated by Tfh, discovery of a Tfh-independent mechanism that enhances humoral immunity during pregnancy is warranted. We propose to study B cell and antibody responses to placental versus foreign antigen. Project hypothesis: Sialoglycan modification of placental antigen, suppressed effector capacity of anti- placental antibodies, and alteration in Tfh-B cell dynamics lead to protection of the conceptus while preserving immunity to foreign antigen. We submit this application in response to RFA-AI-23-027. Aim 1. Define mechanisms of placental antigen-specific B cell suppression (Checkpoint 1). Aim 2. Characterize the non- pathogenic nature of anti-placental antibodies (Checkpoint 2). Aim 3. Evaluate maternal B cell and follicular helper T cell dynamics in response to foreign antigen. Identifying molecular mechanisms suppressing B cell activation (Aim 1) and mapping out tolerance pathways that protect the conceptus from antibody-mediated damage (Aim 2) are the first steps in evaluating how alterations affect pregnancy outcomes. Identification of a pregnancy factor capable of boosting humoral immunity (Aim 3) would be a significant divergence from the textbook paradigm and could become clinically useful to boost responses to vaccines and fight infections.