Empowering Molecular Tumor Boards

Modern oncology is a complex and interdisciplinary practice, often requiring multiple providers from various specialties to collaboratively examine and communicate about patient information from diverse and disparate data sources such as electronic health records (EHR), radiological and pathological imaging, genomics, and other complex molecular data types. The adoption of molecular therapies such as targeted tyrosine kinase inhibitors and immunotherapeutic agents has only increased the amount of information that providers need to consider when deciding how best to treat a patient, as data from all of these disparate data sources informs decisions about treatment.

When standard-of-care therapies are either inadvisable or ineffective, providers often also must navigate the complicated landscape of clinical trials to find patients alternative treatments. Meanwhile, providers such as clinical trial coordinators, data analysts and recruiters face challenges, such as finding suitable patients for trial recruitment and tracking treatment outcomes, in the same complex, multi-modal data at the scale of hundreds or thousands of patients.

Because of the diversity of data types and sources involved in modern oncology, an integrative solution to viewing and analyzing oncology-relevant patient data across data types and provider specialties has historically been unavailable. Providers are often forced to cross-reference data across an EHR system, a radiology viewer, a pathology viewing solution, and various genomic and molecular reports generated by multiple vendors, often with little standardization in terms of the presentation of data. This leads to substantial inefficiencies for providers and hospital systems that can impact the quality of patient care.

For example, scheduling and preparation of complex cases for discussion at tumor boards is a burdensome task for primary oncologists, imaging and genomics specialists, clinical trial coordinators and administrative staff. Each specialist involved in a case often spends many hours compiling and preparing the relevant information for presentation, limiting the number of cases that can be prepared for any given tumor board. The diversity of data types and presentations can lead to inefficiencies in presentation and confusion for providers who do not specialize in a particular area, requiring a tumor board more time to consider any given case.

Furthermore, quantifying the outcomes of tumor board review and subsequent treatment is itself a challenge without an integrated solution to track patient data after the tumor board has considered the case. All of these challenges ultimately mean fewer patients actually receive tumor board review as compared to the total number of patients that could benefit; this type of outcome is echoed across all levels of oncology care, from the individual provider to entire hospital networks, and due at least in part due to the lack of a coherent solution to managing complex multi-modal oncology data.

Solution

To address these challenges, datma has developed the datma.360 patient data visualization and analysis environment with oncology-focused capabilities, including automated clinical trial searching, cohort discovery tools, and integrated scheduling and outcome tracking for tumor boards. By providing a standardized interface to complex multi-modal data, datma.360 enables individual providers to quickly understand the entirety of a patient’s case and to easily collaborate with other providers of multiple specialties in synchronous and distributed formats such as tumor boards.

In addition, data analysts and clinical trial coordinators can utilize scheduling, cohort searches and outcome tracking to enable assist in patient recruitment and in tracking the efficacy of treatments. datma.360 is built on top of the datma.BASE multi-modal data storage and analytics platform, a scalable solution that integrates omics (genomics, transcriptomics), imaging and EHR data into a single coherent representation for visualization and manipulation.

datma.BASE automates the process of importing patient information through customizable ingestion pipelines and integration with industry-standard protocols such as Fast Healthcare Interoperability Resources (FHIR), enabling a standardized and automated ingestion process that relieves much of the data management burden on providers, allowing them to focus on improved patient outcomes.

The datma.360 and datma.BASE solution has been successfully deployed with several organizations, notably a large multi-state hospital network which uses datma.360 as its primary solution for distributed molecular tumor boards and pathology data storage and visualization.

datma.BASE

datma.BASE consists of three primary data storage layers, an automated data ingestion system, and a RESTful API layer that provides access to the underlying data and ingestion system. OmicsDS is designed to handle almost any type of “omics” data (e.g. genomics, transcriptomics, copy number variation, methylation, etc.) and CLIA-compliant storage for customer-provided annotation data. ImageDS is optimized for medical imaging, with a particular focus on pathology imaging. These proprietary data stores are integrated with a database that stores EHR data and documents and coordinates patient information across the data types; this database also stores information for additional modules such as scheduling.

All the components of datma.BASE are securely containerized for simplified deployment in a customer tenancy in a cloud-based, on-premises or hybrid model.

Ingestion of the various data types stored in datma.BASE is accomplished through a flexible ingestion pipeline that handles cross-modality mapping to ensure data across all the data types is consistently mapped to the correct patient. This mapping process also allows for integration with other data sources, such as existing EHR systems using FHIR interfaces. Ingestion is configurable to accommodate standard data formats from multiple vendors and can be extended with customized modules to handle customer-specific proprietary data streams. Logging and automation processes are integrated into the ingestion pipeline to enable automation, reproducibility and auditing.

datma.360

The datma.360 data visualization environment is a web-based visualization solution for multi-modal patient data from single-patient to population-scale. As with datma.BASE, all components of datma.360 are containerized for simplified deployment, and datma.360 itself is designed as a modular system to incorporate customizable functionality for particular customers and use cases.

For oncology-focused use, datma.360 incorporates a unified visualization interface centered around a timeline-based, customizable visualization of treatment, biomarker, genomics, and imaging data and events within a patient’s history. A proprietary pathology visualization system connects to ImageDS and integration with DICOM-based viewers for radiology data is also available. Genomics data is visualized in an intuitive single or multi-sample view with customized annotation for consequences, available treatments and clinical trials.

These interfaces can be used by individual clinicians as a general oncology data visualization tool or employed in synchronous collaborations (e.g. tumor boards), and have been specifically optimized for use in distributed settings using video conferencing.

Scheduling and Coordination

datma.360 incorporates additional modules enabling tumor board-specific scheduling and outcome tracking functionality. The scheduling interface allows administrators to easily schedule patients for a specific tumor board and to inform participating providers through integration with calendar systems such as Microsoft Teams. Facilitators can utilize the scheduling interface to quickly move from patient to patient during a synchronous tumor board session.

The outcome tracking interface allows tumor board facilitators to record specific information about a given patient’s discussion such as particular topics of discussion and if any recommendations were made; this data can later be utilized by analysts and administrators for purposes of auditing, accreditation and outcomes analysis. Outcome tracking is also useful for providers and facilitators to track recommendations and outcomes from previous tumor board sessions with the same patient that may inform the current tumor board’s interpretations and recommendation.

Tumor Board Use Case

datma has deployed datma.360 and datma.BASE at a major multi-state hospital network, where the system is in production use as the platform for the system’s molecular tumor boards, as well as the primary digital pathology imaging storage and visualization solution. In the molecular tumor board setting, datma.BASE is integrated with the hospital system’s existing EHR (via FHIR), genomics and imaging pipelines to ingest data for all in-house molecularly characterized cases, and supports ingestion of data from externa vendors for cases that were sequenced and/or imaged outside of the health system.

datma.360 is also integrated with the health system’s custom clinical annotation pipelines and presents these results to providers during tumor board sessions, while datma.BASE supports both the tumor board application and system-wide cohort searches for clinical trial feasibility based on both clinical and molecular inclusion/exclusion criteria.

Provider response to datma.360 has been enthusiastic and engaged, with several product improvements directly resulting from provider feedback over the deployment cycle. Up to 50 providers have attended biweekly synchronous tumor board sessions from all over the geographic area served by the hospital network, including in underserved and remote areas from which access to tumor boards would traditionally be difficult. Efficiency gains from automated ingestion minimizing preparation time for cases and from a standardized and intuitive interface to the data have also been apparent, enabling consideration of up to six cases per hour-long session.

Finally, enabling integrated access to treatment and clinical trial information has assisted providers in locating suitable trials for many patients and in making appropriate treatment decisions based on molecular and imaging phenotypes. The deployment has been considered a success by the hospital network, and datma is expected to roll out datma.360 and its scheduling interface as a standard for several general tumor boards system-wide in the near future.

Summary

Providers are challenged by the sheer breadth and depth of the information required to make good treatment decisions in modern oncology practice, and the lack of an integrated solution for managing, presenting and analyzing multi-modal data presents barriers to both providers and to improved patient outcomes.

By providing a coherent and integrated data management solution and interface, datma.360 and datma.BASE are proven to eliminate many of these barriers in real-world clinical deployments, reducing the cognitive and data management workload for providers while simultaneously providing innovative opportunities for collaboration in synchronous and asynchronous distributed settings.

By leveraging the technologies of datma.360 and datma.BASE providers can better understand complex oncology data at patient and population scale and take better advantage of the promise of personalized medicine to provide better outcomes for more patients than ever before.