Coremine Vitae - to identify personalized treatment options for cancer patients

Standard cancer therapy with surgery, radiation, and chemotherapy can be described as a "one size fits all" approach. These treatment modalities have been and still are dominating cancer care but are increasingly being replaced by a "one patient, one therapy" approach to cancer care that is based on the individual characteristics of each patient and his / her molecular profile. Immunotherapy with immune checkpoint inhibitors and other targeted therapies are part of what can be called precision cancer medicine (PCM). Immune checkpoint inhibitors and kinase inhibitors have been demonstrated to provide significant clinical benefit for select patients. However, PCM is not yet available to everyone. It is crucial to develop sophisticated bioinformatics tools to efficiently analyze mutation profiles and efficiently identify the best possible treatment options. In response, we want to develop Coremine Vitae (CMV) as a personalized biomedical intelligence solution to identify the best possible treatment option for patients. While the current CMV solution handles many steps manually, we aim to automate the workflow by developing machine learning, big data analytics, and specialized algorithms to identify and rank evidence and treatment options more efficiently. We have designed a data model to store and transfer clinical information between CheckWare and Coremine Vitae software. The data model is general and can be adapted to enable integration with EMR systems. We are looking into the possibility of creating a solution built upon FHIR for allowing direct integration with EMR systems like DIPS and Epic (Helseplattformen). As a part of the automation process, we have developed a framework for digitizing the Coremine Vitae reports for easy utilization of information elements from previous reports. The first sets of AI algorithms based on rule-based reasoning have been developed for creating automated search strategies based on patient profile summary (diagnosis, biomarkers, demographics, treatment history and comorbidities). Algorithms for automated annotation of medically relevant concepts - disease, biomarkers, treatment modalities and drugs - are currently being tested. We have generated and presented more than 10 patient case reports in addition to several mini-review and demo-reports. Together with our partners and external collaborators we have conducted several summative tests to better understand the patient, clinician, and researcher experiences. Results from these tests will further guide the technology development.

-

State-of-the-art treatment of advanced and metastatic cancer is increasingly incorporating next-generation sequencing to characterize tumor DNA and select targeted treatments based on biomarkers such as mutations or other genomic aberrations. Despite fact that Precision Cancer Medicine based on aberration-matched therapies has been demonstrated to provide significant clinical benefits, this is not universally available to most cancer patients. Many cancer drugs are ineffective for a large proportion of patients, 75% or more, consequently, millions of cancer patients are receiving treatments that give little more than unwanted and severe side effects. In response to this, we want to develop Coremine Vitae, a biomedical intelligence system to identify the best possible treatment options to individual patients. The optimal personalized treatment strategy requires consideration of NGS analyses, other biomarkers, patient status, medical history, as well as patient preferences (side effects, mobility, etc.). While the present Coremine Vitae analysis process handles many steps manually, we aim to develop machine learning and big data analytics to identify and rank evidence and treatment options, and thus dramatically decrease the need for manual effort. In addition, the use of the system will improve process results by allowing more comprehensive searches and reduce the risk of bias. Through our partners, Haukeland University Hospital and Sykehuset Innlandet, we have access to patients needed to develop our system, and at the same time obtain clinical validation of the algorithms and resulting Coremine Vitae analysis reports. A successful outcome of this project will allow us to offer Coremine Vitae as a Software as a Service and scale our business substantially enabling us to enter the global market. This will considerably increase the number of patients who can benefit from clinically validated personalized treatment options and anticipated drug response.