Approaches to Addressing Barriers in Adopting Genomic Testing and Targeted Therapy in Oncology
To effectively utilize genomic testing and targeted therapies, the oncologist needs to correlate a patient's demographics and co-morbidities, evidence of cancer as observed in radiology studies and cause of cancer as observed in pathology and genomic tests.
This requires a data management system that can integrate, harmonize, store and facilitate access to multimodel data spanning structured clinical data and notes from EHR, images and report from Radiology systems, images and reports from pathology systems and genomic test results with ability to connect and retrieve information from external knowledge sources for guidelines, clinical trials and therapies.
This data management layer should also provide interfaces for establishing connectivity with the different source systems including FHIR (Fast Healthcare Interoperability Resource) for EHR integrations, DICOM (Digital Imaging and Communications in Medicine) for PACS (picture archiving and communication system) integrations, VCF (Variant Call Format) and external NGS lab formats for genomic to minimize the burden on the information technology teams in the health systems.
In the community oncology setting the radiology, pathology and genomic services are not always performed within the same practice or organization. The data management system should provide the ability for external service providers to easily upload the radiology, pathology and genomic reports and or images, which can then be integrated into the patients clinical timeline without requiring manual intervention.
The data management layer should be augmented with a user interface and visualization that makes it intuitive to navigate a patient's history and understand the interplay between the treatments and disease progression.
The user interface needs to facilitate the following to reduce the cognitive overhead for an oncologist when reviewing a patient's case.
Ability to organize and traverse the longitudinal history of the patient while accounting for temporal nature of the diagnostics, procedures and treatments
Ability to zoom into specific diagnostic events in a clinical timeline and view the reports and images associated with radiology and pathology studies
Annotation of images to enable oncologist to asynchronously interact with the pathologist and radiologist
Comparison of genomic testing results; across time and across panels /labs
Integrated views of somatic and germline test results
Plot biomarkers and lab values against treatment timelines to gain line of sight into patient's response to treatment
The user interface should be easily accessible from the EHR without requiring additional credentials and interference with the clinical workflows.
To enable oncologists to confidently leverage genomic test results in treatment planning decisions they need access to high confidence - evidence based treatment recommendations in a timely manner.
Information needs to be available for the oncologist to leverage during their interaction with the patient. The treatment recommendations should be presented to facilitate the following when evaluating treatment decisions for the patient.
Standard of care guidelines review. What are the guidelines for diagnostics and therapies for a patient who is diagnosed with a specific type of cancer. Being able to view the guidelines that are applicable to a patient without having to manually search for them will make it efficient for oncologist to incorporate the guidelines into their decision making process.
Real World Evidence based evaluation. What diagnostics and treatment options were pursued for “patients like mine”. This would consist of identifying a cohort of similar patients based on their genomic profile and other characteristics, organizing/grouping them by their journey from diagnosis to treatment(s), and comparing the outcomes across the groups. Deviations from the standard of care guidelines should then be further substantiated by peer reviewed publications and data which would help build confidence in selecting these options for treatment.
Efficient access to Virtual Tumor Boards (VTB). While VTB’s having scaling challenges in terms of how many patients can be reviewed in a single session and how frequently these sessions can be held, they still serve as the primary avenue through which oncologists can tap into the expertise of their colleagues within their organization. Providing an efficient method to request case reviews while eliminating the manual steps involved in curating the data for the case review would enable oncologists to tap into the knowledge within their organization in a timely manner. Archiving the VTB discussions and surfacing them for similar cases is another approach to diffuse knowledge and make the decision making process timely.
In addition to the information that is focused on helping the oncologist make treatment decisions they should also be provided with ready access to patient education materials that highlight the benefits and risks of genomic testing and associated therapies. This would enable oncologists to better guide the patients through a combined decision making process.
In order to help oncologists keep up with rapidly changing clinical trials and therapies landscape an “Always-On” clinical trials and therapies matching capability should be provided along with the genomic test results for a patient.
The matching capability should be able to query the repositories of clinical trials like clinicatrials.gov and return the matching trials. The systems should also be able to tap into repositories like OncoKB to identify FDA approved therapies that are appropriate for treating the variants observed in the patients test results.
In addition to the variants observed in the patient’s test results, the system should be able to leverage demographics, co-morbidities, cancer diagnosis, previous treatments, procedures, clinical staging etc from the discrete fields and note in the EHR to ensure the inclusion/exclusion criteria for the clinical trials are accounted for when returning match results.
The match results should be refreshed each time a patient record is accessed by the oncologist thereby ensuring the latest trials and therapies are always being surfaced to the oncologist without them having to search for it. Synchronizing alerts on new matches with patient visits scheduled in the EHR would provide efficiencies in ensuring the oncologist is aware of the latest matches. The oncologist should also be able to align the search with patient preferences like travel distance, phase of the trial, time commitment etc.
While there are regulatory changes that are being put forth to address the Prior Authorization (PA) challenges, the oncologist and the prior authorization teams supporting the oncologist would greatly benefit from automation that would eliminate the most time consuming activities in the current prior authorization process.
Based on the payer, specific templates can be created to capture the relevant information required for validating if a patient is eligible for a specific diagnostic or therapy. The oncologist should be able to trigger the validation which will prompt the system to search the discrete fields and notes in the EHR and bring back evidence to confirm the patient's eligibility for PA.
An example of this scenario is when PA for an NGS test requires the patient to be diagnosed with Stage IV (Metastatic) cancer. The system should be able to find the documentation of the stage in clinician notes or pathology notes and present it to the PA team and thereby cut down the time it takes to manually search and validate this information. In addition the system should be able to pull together the patient's profile and clinical history, supporting evidence for the efficacy and necessity of the requested service for the oncologist and the PA team to review there by expediting the PA request.
Conclusion
Oncologists face considerable challenges in adopting genomic testing and targeted therapies as standard practice for the patients. The challenges include cumbersome user interfaces and siloed nature of the required data, lack of confidence in utilizing genomic and high administrative burden associated with utilizing genomic testing and targeted therapies.
An integrated multimodal data stack that facilitates integration of relevant data with limited IT resource commitment, provides the foundation for leveraging the latest advances in artificial intelligence to automate critical operational tasks and curate relevant real world evidence. Packaging these automations and insights into an highly efficient user experience that can be easily accessed from existing clinical workflow and systems can go a long way in helping oncologists navigate these barriers and provide better outcomes for their patients.
