Patient Genomic Data Research Articles

CANCER DRUG SENSITIVITY THROUGH GENOMIC DATA: INTEGRATING INSIGHTS FOR PERSONALIZED MEDICINE IN THE USA HEALTHCARE SYSTEM

Despite the significant progress in cancer genomics in America, there is still a noteworthy gap regarding genomic markers that predict drug sensitivity which presents a major obstacle to personalized oncology care. Tumors This research project aims to identify a set of genetic variations or mutations that influence the individual response of a cancer patient to certain drugs. This study also aims to develop machine learning models that can analyze a patient's genomic data to predict their likely response to different therapies. This study utilized the Genomic of Drug Sensitivity in Cancer (GDSC). The GDSC dataset is a very valued resource in therapeutic biomarker discovery in cancer research. This dataset combined drug response data with genomic profiles of cancer cell lines, enabling investigations into the relationship between genetic features and drug sensitivity. The main task associated with this dataset was to predict drug sensitivity, measured as IC50 values, from genomic features of cancer cell lines. Several accredited and proven Machine Learning algorithms were utilized in the study, particularly, Linear Regression, Ridge Regression, and SGD Regression. The most important regression model evaluation metrics deployed in a drug sensitivity prediction included the Mean Squared Error- MSE, Root Mean Squared Error- RMSE, and Mean Absolute Error- MAE. The Ridge Regression model outperformed the Linear Regression and the SGD algorithm, particularly, the Ridge Regression model captured excellently the hidden trends in the data much better compared to the other two models. Predictive analytics can significantly enhance clinical decision-making in the USA by providing health professionals with data-driven insights into the best available treatment options. As patient complexity and treatment options continue to grow, such models will help clinicians choose the most appropriate interventions for individual patients, informed by historical data on their disease course and other individual patient factors, including genetic profiling and comorbid status.

Genetic, Metabolic, and Proteomic Polymorphisms and Clinical Phenotypes of Sepsis

The heterogeneity of sepsis patient populations remains an unresolved issue, hindering the development of effective therapeutic strategies and disease prognostic tools. Classification of diverse sepsis patients by molecular endotypes, together with multi-omics profiling, enables a more personalized treatment approach. Studying the immune response, genomic, metabolomic and proteomic profiles of sepsis patients will enable clinical phenotyping of this diverse population and the development of a precision approach to the diagnosis, prognosis and treatment of sepsis and septic shock. The aim of the review was to discuss sepsis subtypes as identified by profiling of patient genomic, metabolic, and proteomic data and present the latest approaches addressing the heterogeneity of sepsis patient populations, such as multi-omics endotyping and clinical phenotyping, which may aid in targeted therapy and optimization of diagnosis and therapy. The keywords «sepsis omics», «sepsis endotypes», and «sepsis heterogeneity» were used to search PubMed databases without language restrictions. From over 300 sources, 120 were selected for analysis as being most relevant to the aim of the review. More than half of these were published within the last five years. Criteria for excluding sources were their inconsistency with the aims of the review and their low informativeness. This review discusses the different types of immune responses, the impact of patient population heterogeneity on therapeutic interventions, and current perspectives on phenotyping sepsis patients. Despite the limitations of centralized collection of clinical information, cluster analysis of large data sets and the role of immune response genomics, metabolomics, and proteomics are beginning to dominate the prognosis and treatment of sepsis. Establishing links between all these elements and attempting clinical phenotyping of sepsis, including subtype analysis, appear to be critical in the search for personalized treatment approaches in the near future. Conclusion. Currently, the widely accepted goal in sepsis management is early detection and initiation of therapy to prevent the development of irreversible septic shock and multiorgan failure syndrome. Personalized genetic, metabolomic and proteomic profiling of the patient seems to be an intriguing and promising avenue in the search for new treatment strategies in sepsis.

YAP1 Status Defines Two Intrinsic Subtypes of LCNEC with Distinct Molecular Features and Therapeutic Vulnerabilities.

Large cell neuroendocrine carcinoma (LCNEC) is a high-grade neuroendocrine malignancy that, like small cell lung cancer (SCLC), is associated with the absence of druggable oncogenic drivers and dismal prognosis. In contrast to SCLC, however, there is little evidence to guide optimal treatment strategies, which are often adapted from SCLC and non-small cell lung cancer approaches. To better define the biology of LCNEC, we analyzed cell line and patient genomic data and performed IHC and single-cell RNA sequencing of core needle biopsies from patients with LCNEC and preclinical models. In this study, we demonstrate that the presence or absence of YAP1 distinguishes two subsets of LCNEC. The YAP1-high subset is mesenchymal and inflamed and is characterized, alongside TP53 mutations, by co-occurring alterations in CDKN2A/B and SMARCA4. Therapeutically, the YAP1-high subset demonstrates vulnerability to MEK- and AXL-targeting strategies, including a novel preclinical AXL chimeric antigen receptor-expressing T cell. Meanwhile, the YAP1-low subset is epithelial and immune-cold and more commonly features TP53 and RB1 co-mutations, similar to those observed in pure SCLC. Notably, the YAP1-low subset is also characterized by the expression of SCLC subtype-defining transcription factors, especially ASCL1 and NEUROD1, and as expected, given its transcriptional similarities to SCLC, exhibits putative vulnerabilities reminiscent of SCLC, including delta-like ligand 3 and CD56 targeting, as is with novel preclinical delta-like ligand 3 and CD56 chimeric antigen receptor-expressing T cells, and DNA damage repair inhibition. YAP1 defines distinct subsets of LCNEC with unique biology. These findings highlight the potential for YAP1 to guide personalized treatment strategies for LCNEC.

Enhanced computed tomography radiomics predicts solute carrier family 7, member 11 in head and neck squamous cell carcinoma.

Traditional prognostic indicators for head and neck squamous cell carcinoma (HNSCC), such as clinicopathological features, human papillomavirus status, and imaging examinations, often lack precision in guiding medical therapy. Therefore, discovering novel tumor biomarkers that can accurately assess prognosis and aid in personalized medical treatment for HNSCC is critical. Solute carrier family 7, member 11 (SLC7A11), is implicated in ferroptosis, and various malignant tumor therapies regulate its expression. However, the mechanisms regulating SLC7A11 expression, the transporter activity, and its specific role in controlling ferroptosis in cancer cells remain unknown. Thus, in this study, we aimed to develop an improved computed tomography (CT) radiomics model that could predict SLC7A11 expression in patients with HNSCC. We used patient genomic data and corresponding augmented CT images for prognostic analysis and building models. Further, we investigated the potential molecular mechanisms underlying SLC7A11 expression in the immune microenvironment. Our radiomics model successfully predicted SLC7A11 mRNA expression in HNSCC tissues and elucidated its association with relevant genes and prognostic outcomes. SLC7A11 expression level was high within tumor tissues and was connected to the infiltration of eosinophil, CD8+ T-cell, and macrophages, which was associated with poor overall survival. Our models demonstrated robust predictive power. The distribution of radiomics scores (RAD scores) within the training and validation sets was markedly different between the high- and low-expression groups of SLC7A11. SLC7A11 is likely an important factor in the prognosis of HNSCC. SLC7A11 expression can be predicted effectively and reliably by radiomics models based on enhanced CT.

CGP-based IO biomarker assessment and treatment utilization across a major US health system.

e23109 Background: Immune checkpoint inhibitors (IO) have become a powerful precision therapy option to treat advanced stage cancer with biomarkers such as PD-L1, tumor mutational burden (TMB), and microsatellite instability (MSI) associated with improved patient response rates. Despite this, many patients do not receive genomic testing for all IO biomarkers. Providence, a large US community health system, developed a pathologist-directed testing protocol where comprehensive genomic profiling (CGP) was routinely used at time of diagnosis for advanced cancer patients. Methods: Advanced cancer patients who received CGP (ProvSeq 523) and IHC testing for PD-L1 between 2019-2023 were included in the study. Patients were required to be treated at Providence and were assessed for presence of IO biomarkers and subsequent therapy selection. Real world data were curated from patient charts and genomic laboratory data by employing a novel natural-language processing (NLP) approach to accelerate abstraction. Results: The study included 2,502 patients (53% female, median age 68y, 83% white). Top 3 tumor types tested were lung (40%), colorectal (9%), and breast (8%). Overall, 58% (N = 1,455) of patients had presence of ≥1 IO biomarker, with 46% (N = 1,155) being PD-L1 positive, and 27% (N = 682) being TMB-H. In PD-L1 negative patients (N = 1,347; 54%), 300 (22%) were TMB-H and 18 (1%) were MSI-H. 53% (N = 767) of patients who harbored an actionable IO biomarker received IO-based therapy. Fewer patients with an IO biomarker received chemotherapy compared to patients without an IO biomarker (23% vs 35%, p < 0.001). Patients who possessed ≥2 IO biomarkers received an IO precision therapy 68% of the time vs 47% in patients with 1 IO biomarker (p < 0.001). In PD-L1 negative patients, 26% (N = 78) of TMB-H patients received IO monotherapy compared to 5% (N = 49) of TMB-L patients. Conclusions: IO biomarker presence is associated with increased precision therapy use. CGP identified many TMB-H patients for IO monotherapy that would have been missed with PD-L1 testing alone. More than half of patients eligible for IO therapy don’t receive it; ongoing analyses will evaluate the impact of pathology-directed reflex testing on IO as well as targeted therapy approaches. [Table: see text]

Abstract NG03: Functional and computational approaches to uncover selection advantages of cancer aneuploidy

Abstract Aneuploidy, including the gain or loss of whole chromosomes or chromosome arms, is a near-universal feature of cancer. We previously applied methods that define chromosome arm aneuploidy to over 10,000 tumors in the Cancer Genome Atlas (TCGA). Cancer subtypes are often characterized by tumor specific patterns of chromosome arm copy number alterations and breakpoints; for example, squamous cell carcinomas (SCCs) from different tissues of origin are characterized by chromosome 3p (chr3p) loss and chromosome 3q (chr3q) gain. From the TCGA aneuploidy data, we developed an algorithm called BISCUT to distinguish peak regions of aneuploidy breakpoints on each chromosome arm. BISCUT identified loci affected by broad copy number alterations that provide fitness advantages or disadvantages both within individual cancer types and across cancers. Our analyses are consistent with selection being the primary driver of aneuploidy events in cancer. We next wanted to validate some BISCUT peaks without a known driver, to identify potential gene deletions that are beneficial in cancer cells. We focused on chromosome 8p (chr8p), as this arm is frequently deleted across cancer types, but no strong tumor suppressors have been identified. Recent advances in genome engineering allow generation of large chromosomal alterations and validation of findings from patient genomic data. For this study, we used our CRISPR-Cas9 arm-deletion system to delete chr8p in human immortalized epithelial cells. Cells with chr8p deletion showed lower amounts of cell death in culture. Knockdown of WRN, one of the two genes in the smallest chr8p BISCUT peak, was sufficient to reproduce this phenotype, suggesting that WRN haploinsufficiency may be beneficial to tumor development. As aneuploidy occurs in tissue-specific patterns, we also wanted to explore the role of chr3 arm aneuploidies in squamous cancers. We again used the CRISPR-Cas9 system to delete one copy of chr3p in a human immortalized lung epithelial cell line similar to the putative cell-of-origin in lung SCC. Consistent with patient data, expression of chr3p genes was decreased upon deletion, as well as increased expression of interferon response genes. Cells with chr3p deletion initially proliferated more slowly than their siblings. Interestingly, after several passages in culture, this proliferation defect was rescued in chr3p deleted cells. Genome sequencing and karyotype analyses suggested that this was partially the result of chr3 duplication, with cells transitioning to a state of chr3q gain. Culturing these cells as organoids also demonstrated that chr3p deletion and chr3q gain can affect basal cell differentiation, consistent with their high frequency in squamous cancers. Our genome engineering approach to model chromosome arm aneuploidies provides a robust model to validate drivers from aneuploidy events, which will be critical to address the gap in our understanding of aneuploidy in cancer. In addition, our methods have identified consequences of individual aneuploidy events, which will lead to new precision oncology targets for patients. Citation Format: Alison M. Taylor. Functional and computational approaches to uncover selection advantages of cancer aneuploidy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr NG03.

Abstract 6455: Investigating clinical trial eligibility criteria to improve MatchMiner trial matching

Abstract As the number of precision medicine (PM) trials and the volume of patient genomic data have grown, it has become challenging for clinicians and trial staff to identify PM trial options for patients. At Dana-Farber Cancer Institute (DFCI), we have addressed this challenge by developing our own open source institutional trial matching software called MatchMiner. MatchMiner algorithmically matches patient genomic and clinical data with PM trial eligibility data. PM trial eligibility data is manually curated into a human- and computer-readable structured format called clinical trial markup language (CTML), for trial matching with DFCI’s institutional genomic assay OncoPanel. MatchMiner has 2 main modes of clinical use: (1) patient-centric, where clinicians can search for trial matches for individual patients and (2) trial-centric, where trial staff can identify patients that match their trial’s genomic eligibility. Thus far, more than 300 DFCI patients have enrolled onto trials facilitated by MatchMiner. In its current implementation, MatchMiner uses limited clinical data (gender, age, and cancer type) to match patients to trials. Although this limited data gives clinicians a head start on assessing available trial options, incorporating more clinical data into MatchMiner would provide more precise trials options and reduce additional work to assess patient eligibility. Here, we investigated clinical trial eligibility criteria that were available for trials at DFCI with the goal of improving specificity of trial matching for MatchMiner. We attempted to answer 2 main research questions 1) what categories of clinical data comprise the eligibility criteria for precision medicine trials at DFCI and 2) which categories are most relevant for trial matching? We first investigated the clinical eligibility criteria across several PM lung cancer trials. After extracting the inclusion and exclusion criteria from the protocol documents, we found criteria could be classified into 10 distinct categories. The most common categories making up almost half of the eligibility criteria were “Prior Therapy” and “Prior or Concurrent Non-Cancerous Disease”. Within some categories, we found it appropriate to also make subcategories to make our analysis more informative such as “Prior Therapy, Surgery” or “Disease Status, CNS”. We next broadened our search to non-lung trials to see if criteria was consistent across other cancer types. We found similar proportions of clinical criteria for the non-lung trials compared to the lung trials, but did find some additional subcategories. Overall, having similar categories of criteria across trials suggests that common clinical data could be used for many different PM trials. Currently, we are investigating which categories are most relevant for trial matching and the use of AI for structuring unstructured clinical data. Citation Format: Harry Klein, Tali Mazor, Matthew Galvin, Jason Hansel, Emily Mallaber, Pavel Trukhanov, James Provencher, James Lindsay, Michael Hassett, Ethan Cerami. Investigating clinical trial eligibility criteria to improve MatchMiner trial matching [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6455.

PillHarmonics: An Orchestrated Pharmacogenetics Medication Clinical Decision Support Service.

Pharmacogenetics (PGx) is increasingly important in individualizing therapeutic management plans, but is often implemented apart from other types of medication clinical decision support (CDS). The lack of integration of PGx into existing CDS may result in incomplete interaction information, which may pose patient safety concerns. We sought to develop a cloud-based orchestrated medication CDS service that integrates PGx with a broad set of drug screening alerts and evaluate it through a clinician utility study. We developed the PillHarmonics service for implementation per the CDS Hooks protocol, algorithmically integrating a wide range of drug interaction knowledge using cloud-based screening services from First Databank (drug-drug/allergy/condition), PharmGKB (drug-gene), and locally curated content (drug-renal/hepatic/race). We performed a user study, presenting 13 clinicians and pharmacists with a prototype of the system's usage in synthetic patient scenarios. We collected feedback via a standard questionnaire and structured interview. Clinician assessment of PillHarmonics via the Technology Acceptance Model questionnaire shows significant evidence of perceived utility. Thematic analysis of structured interviews revealed that aggregated knowledge, concise actionable summaries, and information accessibility were highly valued, and that clinicians would use the service in their practice. Medication safety and optimizing efficacy of therapy regimens remain significant issues. A comprehensive medication CDS system that leverages patient clinical and genomic data to perform a wide range of interaction checking and presents a concise and holistic view of medication knowledge back to the clinician is feasible and perceived as highly valuable for more informed decision-making. Such a system can potentially address many of the challenges identified with current medication-related CDS.

Analysis of Anakinra Therapy for the Deficiency of Interleukin-1 Receptor Antagonist through Clinical Evidence.

Deficiency of interleukin-1 receptor antagonist (DIRA) is a rare life-threatening autosomal recessive autoinflammatory disease with symptoms including but not limited to osteomyelitis, periostitis, and systemic inflammation. DIRA is developed from the loss-of-function biallelic mutations of the IL1RN gene that encodes IL-1 receptor antagonist (IL-1RA), leading to the unchecked pro-inflammatory signaling and subsequent systemic inflammation. Thus, anakinra as the recombinant IL-1RA has become the primary drug to treat DIRA. Although anakinra has been effective for the complete remission of DIRA, it has also shown various side effects. To confirm the efficacy and safety issues associated with DIRA treatment, we conducted a literature review and secondary data analysis to enhance our understanding on this important topic. Through comprehensive literature search, we have identified 15 papers with 25 patients studied. The demographic, clinical, and genetic data were extracted, followed by statistical analysis to support the physiological mechanisms of anakinra treatment. Through the literature review and data analysis, it was found that 88% of patients had complete clinical remission of DIRA upon continual treatment with anakinra; patients had a mean improvement of Hemoglobin (+3.18 g/dL), Erythrocyte Sedimentation Rate (-53.4 mm/h), and C-reactive Protein (-135.45 mg/L) levels, suggesting that the improvement of hematopoietic function and inflammation is a mechanism for anakinra treatment. Various genetic variants were also identified from the patient data that cause the loss of function of IL-1RA, providing real patient genomic data to support the anakinra treatment. Considering the inconsistency and certain variations from clinical research influenced by specific conditions, this review along with the data analysis confirms the efficacy and safety of anakinra treatment for DIRA.

Understanding cancer patient cohorts in virtual reality environment for better clinical decisions: a usability study

BackgroundVisualising patient genomic data in a cohort with embedding data analytics models can provide relevant and sensible patient comparisons to assist a clinician with treatment decisions. As immersive technology is actively used around the medical world, there is a rising demand for an efficient environment that can effectively display genomic data visualisations on immersive devices such as a Virtual Reality (VR) environment. The VR technology will allow clinicians, biologists, and computer scientists to explore a cohort of individual patients within the 3D environment. However, demonstrating the feasibility of the VR prototype needs domain users’ feedback for future user-centred design and a better cognitive model of human–computer interactions. There is limited research work for collecting and integrating domain knowledge into the prototype design.ObjectiveA usability study for the VR prototype–-Virtual Reality to Observe Oncology data Models (VROOM) was implemented. VROOM was designed based on a preliminary study among medical users. The goals of this usability study included establishing a baseline of user experience, validating user performance measures, and identifying potential design improvements that are to be addressed to improve efficiency, functionality, and end-user satisfaction.MethodsThe study was conducted with a group of domain users (10 males, 10 females) with portable VR devices and camera equipment. These domain users included medical users such as clinicians and genetic scientists and computing domain users such as bioinformatics and data analysts. Users were asked to complete routine tasks based on a clinical scenario. Sessions were recorded and analysed to identify potential areas for improvement to the data visual analytics projects in the VR environment. The one-hour usability study included learning VR interaction gestures, running visual analytics tool, and collecting before and after feedback. The feedback was analysed with different methods to measure effectiveness. The statistical method Mann–Whitney U test was used to analyse various task performances among the different participant groups, and multiple data visualisations were created to find insights from questionnaire answers.ResultsThe usability study investigated the feasibility of using VR for genomic data analysis in domain users’ daily work. From the feedback, 65% of the participants, especially clinicians (75% of them), indicated that the VR prototype is potentially helpful for domain users’ daily work but needed more flexibility, such as allowing them to define their features for machine learning part, adding new patient data, and importing their datasets in a better way. We calculated the engaged time for each task and compared them among different user groups. Computing domain users spent 50% more time exploring the algorithms and datasets than medical domain users. Additionally, the medical domain users engaged in the data visual analytics parts (approximately 20%) longer than the computing domain users.

Revolutionizing Breast Healthcare: Harnessing the Role of Artificial Intelligence.

Breast cancer has the highest incidence and second-highest mortality rate among all cancers. The management of breast cancer is being revolutionized by artificial intelligence (AI), which is improving early detection, pathological diagnosis, risk assessment, individualized treatment recommendations, and treatment response prediction. Nuclear medicine has used artificial intelligence (AI) for over 50 years, but more recent advances in machine learning (ML) and deep learning (DL) have given AI in nuclear medicine additional capabilities. AI accurately analyzes breast imaging scans for early detection, minimizing false negatives while offering radiologists reliable, swift image processing assistance. It smoothly fits into radiology workflows, which may result in early treatments and reduced expenditures. In pathological diagnosis, artificial intelligence improves the quality of diagnostic data by ensuring accurate diagnoses, lowering inter-observer variability, speeding up the review process, and identifying errors or poor slides. By taking into consideration nutritional, genetic, and environmental factors, providing individualized risk assessments, and recommending more regular tests for higher-risk patients, AI aids with the risk assessment of breastcancer. The integration of clinical and genetic data into individualized treatment recommendations by AI facilitates collaborative decision-making and resource allocation optimization while also enabling patient progress monitoring, drug interaction consideration, and alignment with clinical guidelines. AI is used to analyze patient data, imaging, genomic data, and pathology reports in order to forecast how a treatment would respond. These models anticipate treatment outcomes, make sure that clinical recommendations are followed, and learn from historical data. The implementation of AI in medicine is hampered by issues with data quality, integration with healthcare IT systems, data protection, bias reduction, and ethical considerations, necessitating transparency and constant surveillance. Protecting patient privacy, resolving biases, maintaining transparency, identifying fault for mistakes, and ensuring fair access are just a few examples of ethical considerations. To preserve patient trust and address the effect on the healthcare workforce, ethical frameworks must be developed. The amazing potential of AI in the treatment of breast cancer calls for careful examination of its ethical and practical implications. We aim to review the comprehensive role of artificial intelligence in breast cancer management.

Tri©DB: an integrated platform of knowledgebase and reporting system for cancer precision medicine

BackgroundWith the development of cancer precision medicine, a huge amount of high-dimensional cancer information has rapidly accumulated regarding gene alterations, diseases, therapeutic interventions and various annotations. The information is highly fragmented across multiple different sources, making it highly challenging to effectively utilize and exchange the information. Therefore, it is essential to create a resource platform containing well-aggregated, carefully mined, and easily accessible data for effective knowledge sharing.MethodsIn this study, we have developed “Consensus Cancer Core” (Tri©DB), a new integrative cancer precision medicine knowledgebase and reporting system by mining and harmonizing multifaceted cancer data sources, and presenting them in a centralized platform with enhanced functionalities for accessibility, annotation and analysis.ResultsThe knowledgebase provides the currently most comprehensive information on cancer precision medicine covering more than 40 annotation entities, many of which are novel and have never been explored previously. Tri©DB offers several unique features: (i) harmonizing the cancer-related information from more than 30 data sources into one integrative platform for easy access; (ii) utilizing a variety of data analysis and graphical tools for enhanced user interaction with the high-dimensional data; (iii) containing a newly developed reporting system for automated annotation and therapy matching for external patient genomic data. Benchmark test indicated that Tri©DB is able to annotate 46% more treatments than two officially recognized resources, oncoKB and MCG. Tri©DB was further shown to have achieved 94.9% concordance with administered treatments in a real clinical trial.ConclusionsThe novel features and rich functionalities of the new platform will facilitate full access to cancer precision medicine data in one single platform and accommodate the needs of a broad range of researchers not only in translational medicine, but also in basic biomedical research. We believe that it will help to promote knowledge sharing in cancer precision medicine. Tri©DB is freely available at www.biomeddb.org, and is hosted on a cutting-edge technology architecture supporting all major browsers and mobile handsets.

Abstract 17773: Novel Partition Scores Identify Pathways Implicated in Clonal Hematopoiesis of Indeterminate Potential (CHIP) Mediated Heart Failure

Background: Peripheral blood somatic genetic variants such as those in clonal hematopoiesis of indeterminate potential (CHIP) increase the risk of both ischemic and non-ischemic cardiomyopathy but mechanisms beyond inflammation are unclear. Research question/Objectives: We sought to understand the prevalence and impact of CHIP and extended somatic variants (ExSV) using a new deep learning algorithm in heart failure patients on the outcome of all-cause mortality and leverage novel partition scores to identify pathways involved. Methods: CHIP and ExSV were evaluated using a novel proprietary deep learning algorithm in a heart failure cohort at Yale using whole exome sequencing and somatic variants identified were validated using UKB and TOPMED cohorts. Cox regression including somatic variants on the outcome of death was performed. Partition score analysis integrating patient demographics, co-morbidities, genomic and transcriptomic data was performed to identify traits and biological pathways implicated in heart failure Results: CHIP was highly prevalent in the 189 patient Yale heart failure cohort at 42% with mean age of 66. Fifty percent had diabetes and 46% had CAD. Using UKB and TOPMED cohorts, novel “CHIP-like” somatic variants were identified that conferred increased or decreased risk of heart failure. Some ExSV were associated with increased risk of death. Partition score assessment identified top traits associated with increased risk of heart failure, including LDL, triglycerides, diabetes, BMI and CAD and pathway analysis identified that metabolic pathways including cholesterol metabolism, glucose signaling and AMPK were implicated in the heart failure phenotype mediated by somatic variants (Table). Conclusions: CHIP and ExSV are prevalent in heart failure patients, impact the outcome of death and metabolic pathways are implicated in mediating the effects of somatic variants/CHIP in heart failure

Definition and Clinical Evaluation for Trimethoprim-Sulfamethoxazole Severe Acute Respiratory Failure.

Trimethoprim-sulfamethoxazole (TMP-SMX)-associated severe acute respiratory distress syndrome (ARDS) has gone underrecognized. We propose the first disease definition and clinical evaluation for a novel adverse drug reaction (ADR) based on a series of recently identified rare cases of life-threatening ADRs. A retrospective study was conducted. All medical records were evaluated. Available pathology samples were sent to Massachusetts General for clinical consultation. Blood samples from surviving patients were obtained and human leukocyte antigen (HLA) analysis was performed by the Children's Mercy Hospital Genomic Center and Vanderbilt University Medical Center. U.S. ICUs, 1996-2021. Nineteen young patients (10-37) were identified. Patients were previously healthy, with no preexisting pulmonary disease, no other cause for respiratory failure, and no chronic history of smoking/vaping. None. Through our retrospective analysis, we analyzed clinical characteristics associated with TMP-SMX. Pathology samples were reviewed, and HLA analysis was performed on available samples by the study team or as standard of care at treatment hospitals in some cases. In 19 critically ill patients, we identified a pattern of severe respiratory failure requiring ICU admission, mechanical ventilation, and frequent extracorporeal membrane oxygenation use. We describe the first three-part clinical diagnosis and evaluation strategy: 1) Clinical definition: Unexplained severe respiratory failure in a patient receiving greater than or equal to 6 days of TMP-SMX at treatment dose (not prophylaxis). TMP-SMX ARDS is a diagnosis of exclusion. 2) Genetic association: One hundred percent of currently available TMP-SMX respiratory failure patient genomic data, ( n = 11) have been carriers of both HLA-B*07:02 and HLA-C*07:02 alleles. HLA allele evaluation could be considered in patients with suspected TMP-SMX respiratory failure. 3) Lung pathology: A unique pulmonary pathologic pattern of lung injury termed diffuse alveolar injury with delayed epithelialization has been observed in these cases. In suspected cases, surgical lung biopsy early in the clinical course could be considered. TMP-SMX is a commonly prescribed antibiotic. However, we find it imperative to share this relatively rare but life-threatening condition with clinicians as the mortality rate approaches 40%.

Unravelling the application of machine learning in cancer biomarker discovery

Machine learning is playing an increasingly important role in the healthcare industry by transforming the way cancer is diagnosed and treated. By analyzing patient data, genomic data, and imaging data, machine learning algorithms can identify molecular signatures that distinguish cancer patients from healthy patients. Biomarkers that can accurately detect and diagnose cancer can be identified through analysis of these data sources. Additionally, personalized cancer therapies can be developed by identifying the most effective treatments based on individual patient characteristics and cancer type. Some of the machine learning techniques used for cancer biomarker discovery include deep learning and support vector machines, which can respectively identify complex patterns in data and classify data to identify relevant biomarkers. The benefits of using machine learning for cancer biomarker discovery are significant, including more precise and personalized treatments, improved patient outcomes, and the potential to transform cancer diagnosis and treatment. However, there are also challenges associated with using machine learning for cancer biomarker discovery, such as data collection and privacy issues, as well as the need for more powerful computational resources. This article explores the potential of machine learning in cancer biomarker discovery and argues that ongoing research in this field has the potential to revolutionize cancer diagnosis and treatment. Future research directions should focus on further developing machine learning algorithms and effective data collection and privacy protocols.

Radiogenomic Associations Clear Cell Renal Cell Carcinoma: An Exploratory Study

Introduction: This study investigates how quantitative texture analysis can be used to non-invasively identify novel radiogenomic correlations with clear cell renal cell carcinoma (ccRCC) biomarkers. Methods: The Cancer Genome Atlas-Kidney Renal Clear Cell Carcinoma open-source database was used to identify 190 sets of patient genomic data that had corresponding multiphase contrast-enhanced CT images in The Cancer Imaging Archive. 2,824 radiomic features spanning fifteen texture families were extracted from CT images using a custom-built MATLAB software package. Robust radiomic features with strong inter-scanner reproducibility were selected. Random forest, AdaBoost, and elastic net machine learning (ML) algorithms evaluated the ability of the selected radiomic features to predict the presence of 12 clinically relevant molecular biomarkers identified from the literature. ML analysis was repeated with cases stratified by stage (I/II vs. III/IV) and grade (1/2 vs. 3/4). 10-fold cross validation was used to evaluate model performance. Results: Before stratification by tumor grade and stage, radiomics predicted the presence of several biomarkers with weak discrimination (AUC 0.60–0.68). Once stratified, radiomics predicted KDM5C, SETD2, PBRM1, and mTOR mutation status with acceptable to excellent predictive discrimination (AUC ranges from 0.70 to 0.86). Conclusions: Radiomic texture analysis can potentially identify a variety of clinically relevant biomarkers in patients with ccRCC and may have a prognostic implication.

Abstract 1067: MatchMiner: An open-source AI precision medicine trial matching platform

Abstract As the number of precision medicine (PM) trials and patient genomic data has grown, it has become challenging for clinicians and trial staff to identify PM trial options for patients. Several trial matching software platforms have been developed to match genomic data from patients with PM trials, but these existing platforms are proprietary and are not easily accessible for adoption by institutions. At Dana-Farber Cancer Institute (DFCI), we have addressed this challenge by developing our own open-source institutional trial matching software, MatchMiner. MatchMiner algorithmically matches patient genomic and clinical data with PM trial eligibility data. Trial eligibility data is manually curated into a human-readable markup language, called clinical trial markup language (CTML), for matching with patient genomic data. MatchMiner has 2 main modes of clinical use: (1) patient-centric, where clinicians search for trial matches for individual patients and (2) trial-centric, where trial staff identify patients that match their trial’s genomic eligibility. We recently described MatchMiner’s usage at DFCI and since our report, we have added 90 additional trial consents facilitated by MatchMiner (>250 trial consents, called MatchMiner consents [MMC]). Here, we describe new characteristics of our MMC including which user mode (patient-centric or trial-centric) was used to match the consent, genomic alterations and cancer types that matched to eligibility criteria, and whether the patient went onto trial. MMCs were mostly identified by patient-centric mode (70%), genomic alterations and cancer types among MMC were diverse (n=55 genes and n=20 cancer types), and 87% of MMC went on trial. Among MMCs, the most common altered genes leading to trial eligibility were ERBB2 and KRAS in breast cancer and lung cancer, which is consistent with the number of therapies targeting ERBB2 and KRAS. MMCs also included patients with rare cancer types, like extraskeletal myxoid chondrosarcoma, as well as rare genomic alterations, such as NTRK fusions. Thus, MatchMiner has been successful at facilitating PM trial matching for a broad range of genomic alterations and cancer types at DFCI. MatchMiner matches patients to trials as soon as their genomic report is available, however, many patients are not yet ready to enroll onto a trial because their cancer is responding to the standard of care or they are in a remission period. To address this problem, we are evaluating the use of artificial intelligence (AI) to identify patients that may be ready for a new treatment option. After trial matches have been generated by MatchMiner, radiology scan text from patients’ tumor scans is run through a natural language processing (NLP) model to identify patients who are more likely to be ready to enroll onto a trial. By using NLP to filter trial matches, we hope to improve MatchMiner’s efficiency of finding trial matches and provide more timely trial options for patients. Citation Format: Harry Klein, Tali Mazor, Matthew Galvin, Jason Hansel, Emily Mallaber, Pavel Trukhanov, Joyce Yu, James Lindsay, Kenneth Kehl, Michael Hassett, Ethan Cerami. MatchMiner: An open-source AI precision medicine trial matching platform [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1067.

Abstract P5-14-08: Machine learning based histopathology images analysis reveals cancer stemness in TNBC patient with 17p loss

Abstract Background: The area of computational pathology has made huge progresses due to advances in artificial intelligence (AI) and machine learning technologies. It has been applied to many research and translational tasks which provide great improvement on medical diagnosis and treatment. Cancer stem-like cells (CSC) have been consistently reported for its key role in Triple negative breast cancer (TNBC) . Given the large amount of existing H&E stained histological slides of TNBC, digital identification of CSC could benefit the evaluation of tumor status and prediction of patients’ response to chemotherapy. Here we proposed an AI framework based on Convolutional Neural Network (CNN) to predict CSC from the histological images of TNBC patient. And our preliminary work suggested that chromosome 17p loss, a common genetic variations in breast cancer, is linked to cancer stemness. Methods: A modified GoogleNet model was adopted as our CNN classifier. Consecutive breast cancer tissue microarrays (TMA), which stained with H&E ,SOX2, OCT4 and NANOG antibodies respectively by IHC, were used as training dataset for the CNN model. Gene expression data from the TCGA and METABRIC datasets were used to identify gene signatures associated with CSC. The connectivity map (CMAP) and Cancer Cell Line Encyclopedia (CCLE) were used for screening compounds that target stemness in cancer cell lines with chromosomal 17p loss. HS578T and EO771 cells with or without heterozygous 17p loss (11b in EO771) were used for in vitro experiments. Female immunodeficient nude (Nu/J) mice were used for animal studies. Results: The well trained GoogleNet model was applied to TNBC patient diagnosis images in TCGA BRCA dataset. By analyzing patient genomic alteration on chromosomal level, we found that loss of chromosome 17p associate with high cancer stemness in TNBC. Flow cytometry assays also demonstrated higher ALDH1 activity and higher CD44+/CD24−/low cell population in HS578T cells with 17p loss. RNA-seq of HS578T cells revealed that most CSC marker genes were located in the unregulated differentially expressed genes (DEGs) of 17p loss cells. We next compared the cytotoxicity of chemotherapy drugs including doxorubicin, paclitaxel, docetaxel and cisplatin on 17p loss and 17p intact HS578T cells, 11b loss and 11b intact EO771 cells, in terms of IC50 value. The IC50 value of indicated drug on 17p loss HS578T cells with were 3-6 fold higher than their IC50 on 17p intact HS578T cells. Similar result was observed in EO771 cells. Next, 17p loss and 17p intact HS578T cells were orthotopically implanted into the Nu/J mice. Under the doxorubicin treatment, mice bearing 17p loss HS578T derived tumors had larger and heavier tumors in compare to mice bearing 17p intact tumors. Next, we did a computational drug screening to identify compounds that can target the cancer stemness in 17p loss cells. Screened out compounds were tested for their cytotoxicity on 17p loss and 17p intact HS578T cells and FK866 showed the most pronounced efficacy on inhibiting the viability of 17p loss cells, compared to 17p intact cells. Followup experiments demonstrated that FK866 can decrease the CSC features induced by doxorubicin, both in vitro and in vivo. FK866 also potentiates the effect of doxorubicin on treating TNBC cells with 17p loss, which provide a drug combination potential for TNBC patient with 17p loss. Conclusions: A CNN based model was developed to identify CSC from TNBC histopathology images. The images analysis combined with patient genomic data revealed that chromosome 17p loss associated with cancer stemness in TNBC. This result was confirmed using assays on TNBC cells with or without 17p loss. A computational drug screening was performed to identify candidates that targeting stemness in 17p loss cells. FK866 was identified and it potentiates the anti-tumor effect of doxorubicin on treating TNBC cells with 17p loss. Our study provides a novel strategy on applying AI to precision treatment for cancers. Citation Format: tianhan dong, jiannan liu, yuanzhang Fang, Ziyu Liu, Xiongbin Lu, kun huang. Machine learning based histopathology images analysis reveals cancer stemness in TNBC patient with 17p loss [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-14-08.

Disruption to the FOXO-PRDM1 axis resulting from deletions of chromosome 6 in acute lymphoblastic leukaemia

A common problem in the study of human malignancy is the elucidation of cancer driver mechanisms associated with recurrent deletion of regions containing multiple genes. Taking B-cell acute lymphoblastic leukaemia (B-ALL) and large deletions of 6q [del(6q)] as a model, we integrated analysis of functional cDNA clone tracking assays with patient genomic and transcriptomic data, to identify the transcription factors FOXO3 and PRDM1 as candidate tumour suppressor genes (TSG). Analysis of cell cycle and transcriptomic changes following overexpression of FOXO3 or PRDM1 indicated that they co-operate to promote cell cycle exit at the pre-B cell stage. FOXO1 abnormalities are absent in B-ALL, but like FOXO3, FOXO1 expression suppressed growth of TCF3::PBX1 and ETV6::RUNX1 B-ALL in-vitro. While both FOXOs induced PRDM1 and other genes contributing to late pre-B cell development, FOXO1 alone induced the key transcription factor, IRF4, and chemokine, CXCR4. CRISPR-Cas9 screening identified FOXO3 as a TSG, while FOXO1 emerged as essential for B-ALL growth. We relate this FOXO3-specific leukaemia-protective role to suppression of glycolysis based on integrated analysis of CRISPR-data and gene sets induced or suppressed by FOXO1 and FOXO3. Pan-FOXO agonist Selinexor induced the glycolysis inhibitor TXNIP and suppressed B-ALL growth at low dose (ID50 < 50 nM).

Balancing the safeguarding of privacy and data sharing: perceptions of genomic professionals on patient genomic data ownership in Australia.

There are inherent complexities and tensions in achieving a responsible balance between safeguarding patients' privacy and sharing genomic data for advancing health and medical science. A growing body of literature suggests establishing patient genomic data ownership, enabled by blockchain technology, as one approach for managing these priorities. We conducted an online survey, applying a mixed methods approach to collect quantitative (using scale questions) and qualitative data (using open-ended questions). We explored the views of 117 genomic professionals (clinical geneticists, genetic counsellors, bioinformaticians, and researchers) towards patient data ownership in Australia. Data analysis revealed most professionals agreed that patients have rights to data ownership. However, there is a need for a clearer understanding of the nature and implications of data ownership in this context as genomic data often is subject to collective ownership (e.g., with family members andlaboratories). This research finds that while the majority of genomic professionals acknowledge the desire for patient data ownership, bioinformaticians and researchers expressed more favourable views than clinical geneticists and genetic counsellors, suggesting that their views on this issue may be shaped by how closely they interact with patients as part of their professional duties. This research also confirms that stronger health system infrastructure is a prerequisite for enabling patient data ownership, which needs to be underpinned by appropriate digital infrastructure (e.g., central vs. decentralised data storage), patient identity ownership (e.g., limited vs. self-sovereign identity), and policy at both federal and state levels.