Prediction of Effectiveness and Toxicities of Immune Checkpoint Inhibitors Using Real-World Patient Data.

PURPOSE: Although immune checkpoint inhibitors (ICIs) have improved outcomes in certain patients with cancer, they can also cause life-threatening immunotoxicities. Predicting immunotoxicity risks alongside response could provide a personalized risk-benefit profile, inform therapeutic decision making, and improve clinical trial cohort selection. We aimed to build a machine learning (ML) framework using routine electronic health record (EHR) data to predict hepatitis, colitis, pneumonitis, and 1-year overall survival. METHODS: Real-world EHR data of more than 2,200 patients treated with ICI through December 31, 2018, were used to develop predictive models. Using a prediction time point of ICI initiation, a 1-year prediction time window was applied to create binary labels for the four outcomes for each patient. Feature engineering involved aggregating laboratory measurements over appropriate time windows (60-365 days). Patients were randomly partitioned into training (80%) and test (20%) sets. Random forest classifiers were developed using a rigorous model development framework. RESULTS: The patient cohort had a median age of 63 years and was 61.8% male. Patients predominantly had melanoma (37.8%), lung cancer (27.3%), or genitourinary cancer (16.4%). They were treated with PD-1 (60.4%), PD-L1 (9.0%), and CTLA-4 (19.7%) ICIs. Our models demonstrate reasonably strong performance, with AUCs of 0.739, 0.729, 0.755, and 0.752 for the pneumonitis, hepatitis, colitis, and 1-year overall survival models, respectively. Each model relies on an outcome-specific feature set, though some features are shared among models. CONCLUSION: To our knowledge, this is the first ML solution that assesses individual ICI risk-benefit profiles based predominantly on routine structured EHR data. As such, use of our ML solution will not require additional data collection or documentation in the clinic.

Landscape Analysis of Breast Cancer and Acute Myeloid Leukemia Trials Using the My Cancer Genome Clinical Trial Data Model.

PURPOSE: The field of oncology is expanding rapidly. New trials are opening as an increasing number of therapeutic agents are being investigated before they can become approved therapies. Aggregate views of these data, particularly data associated with diseases, biomarkers, and drugs, can be helpful in understanding the trends in current research as well as existing gaps in cancer care. METHODS: In this paper, we performed a landscape analysis for breast cancer and acute myeloid leukemia related trials with structured, curated data from clinical trials using the My Cancer Genome clinical trial knowledgebase. RESULTS: We have performed detailed analytics on breast cancer (N = 1,128) and acute myeloid leukemia trial sets (N = 483) to highlight the top biomarkers, drug classes, and drugs-thereby supporting a full view of biomarkers, biomarker groups, and drugs that are currently being explored in these respective diseases. CONCLUSION: Analysis and data visualization of the cancer clinical trial landscape can inform strategic planning for new trial designs and trial activation at a particular site.

My Cancer Genome: Coevolution of Precision Oncology and a Molecular Oncology Knowledgebase.

PURPOSE: The My Cancer Genome (MCG) knowledgebase and resulting website were launched in 2011 with the purpose of guiding clinicians in the application of genomic testing results for treatment of patients with cancer. Both knowledgebase and website were originally developed using a wiki-style approach that relied on manual evidence curation and synthesis of that evidence into cancer-related biomarker, disease, and pathway pages on the website that summarized the literature for a clinical audience. This approach required significant time investment for each page, which limited website scalability as the field advanced. To address this challenge, we designed and used an assertion-based data model that allows the knowledgebase and website to expand with the field of precision oncology. METHODS: Assertions, or computationally accessible cause and effect statements, are both manually curated from primary sources and imported from external databases and stored in a knowledge management system. To generate pages for the MCG website, reusable templates transform assertions into reconfigurable text and visualizations that form the building blocks for automatically updating disease, biomarker, drug, and clinical trial pages. RESULTS: Combining text and graph templates with assertions in our knowledgebase allows generation of web pages that automatically update with our knowledgebase. Automated page generation empowers rapid scaling of the website as assertions with new biomarkers and drugs are added to the knowledgebase. This process has generated more than 9,100 clinical trial pages, 18,100 gene and alteration pages, 900 disease pages, and 2,700 drug pages to date. CONCLUSION: Leveraging both computational and manual curation processes in combination with reusable templates empowers automation and scalability for both the MCG knowledgebase and MCG website.

Framework for Implementing and Tracking a Molecular Tumor Board at a National Cancer Institute-Designated Comprehensive Cancer Center.

BACKGROUND: Over the past few years, tumor next-generation sequencing (NGS) panels have evolved in complexity and have changed from selected gene panels with a handful of genes to larger panels with hundreds of genes, sometimes in combination with paired germline filtering and/or testing. With this move toward increasingly large NGS panels, we have rapidly outgrown the available literature supporting the utility of treatments targeting many reported gene alterations, making it challenging for oncology providers to interpret NGS results and make a therapy recommendation for their patients. METHODS: To support the oncologists at Vanderbilt-Ingram Cancer Center (VICC) in interpreting NGS reports for patient care, we initiated two molecular tumor boards (MTBs)-a VICC-specific institutional board for our patients and a global community MTB open to the larger oncology patient population. Core attendees include oncologists, hematologist, molecular pathologists, cancer geneticists, and cancer genetic counselors. Recommendations generated from MTB were documented in a formal report that was uploaded to our electronic health record system. RESULTS: As of December 2020, we have discussed over 170 patient cases from 77 unique oncology providers from VICC and its affiliate sites, and a total of 58 international patient cases by 25 unique providers from six different countries across the globe. Breast cancer and lung cancer were the most presented diagnoses. CONCLUSION: In this article, we share our learning from the MTB experience and document best practices at our institution. We aim to lay a framework that allows other institutions to recreate MTBs. IMPLICATIONS FOR PRACTICE: With the rapid pace of molecularly driven therapies entering the oncology care spectrum, there is a need to create resources that support timely and accurate interpretation of next-generation sequencing reports to guide treatment decision for patients. Molecular tumor boards (MTB) have been created as a response to this knowledge gap. This report shares implementation strategies and best practices from the Vanderbilt experience of creating an institutional MTB and a virtual global MTB for the larger oncology community. This report describe a reproducible framework that can be adopted to initiate MTBs at other institutions.

Learnings From Precision Clinical Trial Matching for Oncology Patients Who Received NGS Testing.

PURPOSE: Tumor next-generation sequencing reports typically generate trial recommendations for patients based on their diagnosis and genomic profile. However, these require additional refinement and prescreening, which can add to physician burden. We wanted to use human prescreening efforts to efficiently refine these trial options and also elucidate the high-value parameters that have a major impact on efficient trial matching. METHODS: Clinical trial recommendations were generated based on diagnosis and biomarker criteria using an informatics platform and were further refined by manual prescreening. The refined results were then compared with the initial trial recommendations and the reasons for false-positive matches were evaluated. RESULTS: Manual prescreening significantly reduced the number of false positives from the informatics generated trial recommendations, as expected. We found that trial-specific criteria, especially recruiting status for individual trial arms, were a high value parameter and led to the largest number of automated false-positive matches. CONCLUSION: Reflex clinical trial matching approaches that refine trial recommendations based on the clinical details as well as trial-specific criteria have the potential to help alleviate physician burden for selecting the most appropriate trial for their patient. Investing in publicly available resources that capture the recruiting status of a trial at the cohort or arm level would, therefore, allow us to make meaningful contributions to increase the clinical trial enrollments by eliminating false positives.

Stem Cell-Derived Insulin-Producing Cells in 3D Engineered Tissue in a Perfusion Flow Bioreactor.

To circumvent the lack of donor pancreas, insulin-producing cells (IPCs) derived from pluripotent stem cells emerged as a viable cell source for the treatment of type 1 diabetes. While it has been shown that IPCs can be derived from pluripotent stem cells using various protocols, the long-term viability and functional stability of IPCs in vitro remains a challenge. Thus, the principles of three-dimensional (3D) tissue engineering and a perfusion flow bioreactor were used in this study to establish 3D microenvironment suitable for long-term in vitro culture of IPCs-derived from mouse embryonic stem cells. It was observed that in static 3D culture of IPCs, the viability decreased gradually with longer time in culture. However, when a low flow (0.02 mL/min) was continuously applied to 3D IPC containing tissues, enhanced survival and function of IPCs were demonstrated. IPCs cultured under low flow exhibited a significantly enhanced glucose responsiveness and upregulation of Ins1 compared to that of static culture. In summary, this study demonstrates the feasibility and benefits of 3D engineered tissue environment combined with perfusion flow in vitro for culturing stem cell-derived IPCs. Impact statement This in vitro three-dimensional tissue system combined with the flow can be used to better understand the role of biophysical cues that facilitates improved function and maturation of stem cell-derived insulin-producing cells, which can ultimately advance the field of pancreatic tissue engineering as well as in diabetes treatment.

Conceptual Framework to Support Clinical Trial Optimization and End-to-End Enrollment Workflow.

In this work, we present a conceptual framework to support clinical trial optimization and enrollment workflows and review the current state, limitations, and future trends in this space. This framework includes knowledge representation of clinical trials, clinical trial optimization, clinical trial design, enrollment workflows for prospective clinical trial matching, waitlist management, and, finally, evaluation strategies for assessing improvement.