MatchMiner

MatchMiner is an open computational platform for matching patient-specific genomic profiles to precision cancer medicine clinical trials.

7 9

Written by Ethan Cerami

Updated on 14:25, May 20, 2016

Focus of Innovation

Data collection
Data interpretation
Patient selection

Name of applicant

Ethan Cerami

Title of applicant

Director, Knowledge Systems Group

Affiliation of applicant

Dana-Farber Cancer Institute (DFCI)

Your social media links

https://www.linkedin.com/in/ecerami
http://www.twitter.com/ecerami

Names, titles and affiliations of teammates

* James Lindsay, Senior Bioinformatics Engineer, Knowledge Systems Group - Dana-Farber Cancer Institute
* Priti Kumari, Bioinformatics Engineer, Knowledge Systems Group - Dana-Farber Cancer Institute
* Adem Albayrak, Team Lead, Clinical & Translational Informatics - Dana-Farber Cancer Institute
* Bernd van der Veen, Software Developer, The Hyve, Netherlands

MatchMiner is an open source computational platform for matching patient-specific genomic profiles to precision cancer medicine clinical trials. The platform is being developed at Dana-Farber Cancer Institute, where it will be used to enable precision medicine initiatives, based on Profile, our enterprise genomic profiling initiative available to all patients.

The platform is currently being developed in two distinct stages, and piloted at DFCI, after which point the entire platform will be made fully open source, and available to other institutions. Stage 1 of the MatchMiner platform is focused on “trial-centric” matching, enabling clinical trial investigators to create individualized genomic filters, and use these filters to: 1) forecast clinical trial enrollment, based on two-years of genomic sequencing at DFCI; 2) retrospectively identify new patients for clinical trials; and 3) receive alerts of newly sequenced patients matching specific genomic criteria. Stage 2 of the platform is focused on “patient-centric” matching, enabling any clinician to view matching clinical trials for their specific patient, based on genomic eligibility and real-time clinical trial enrollment slot availability. Stage 2 will require central curation of clinical trials at DFCI, and as part of this curation effort, we aim to create an open standard for representing genomic eligibility criteria for clinical trials.

MatchMiner will be launched at DFCI in Q1 2016 as a fully HIPAA compliant platform, and we are currently looking for external partners to extend and pilot the platform.

Why would your idea have a significant impact?

As of 2016, every single cancer center ranked in the top 10 by US News and World Reports currently performs targeted genomic profiling on all or a large number of patients. These initiatives are critical to enabling precision cancer medicine, as molecular information from genomic profiling is currently used to guide treatment decisions and clinical trial enrollment. However the recruitment (or matching) of patients to genomically driven clinical trials is often a manual process, resulting in overall inefficiencies and missed opportunities. The MatchMiner platform aims to aid clinicians in automating this process, with the overall goal of increasing clinical trial enrollment in precision cancer medicine clinical trials, and maximizing clinical trial options for all cancer patients.

What are the biggest hurdles to implementing your idea?

At DFCI and other major cancer centers, our biggest hurdle will be adoption by clinicians. Clinicians are faced with huge demands on their time, and already interact with dozens of information systems, including electronic medical records (EMR) and pathology information systems. We therefore face a significant challenge in creating a useful, user-friendly system that works within the existing workflow of clinical trial investigators and other working oncologists. Beyond DFCI, adoption of MatchMiner across multiple cancer centers will face two additional challenges: 1) each cancer center has a unique genomic profiling pipeline, tied to a unique system for pathology and clinical interpretation; each cancer center also has a unique set of information systems for managing clinical trials. Creating an open-source system capable of spanning the diversity of genomic profiling and the diversity of information systems within cancer centers will be a significant challenge; 2) there currently exists no structured standards for representing clinical trial eligibility, representing a significant hurdle for interpreting clinical trials from existing resources, such as http://clinicaltrials.gov.

How difficult will it be to overcome those obstacles?

We aim to overcome these specific obstacles in multiple ways. First, we have engaged clinicians in the earliest possible stages of MatchMiner development, and have conducted usability testing with early clinical adopters. Second, we have made plans to open source MatchMiner, and work with other cancer centers to adopt and extend the platform. Third, we aim to create an open standard for representing genomic clinical trial eligibility, easing the process of sharing clinical trial enrollment across multiple cancer centers.

How would you test for the impact of this idea and how would you quantify the impact?

The impact of MatchMiner will initially be quantified in collaboration with several groups managing phase 1 clinical trials at DFCI. Two key metrics will be observed; enrollment rate and number of patient reviews an oncologist must perform to reach enrollment goals. These two factors greatly influence the efficiency of the recruitment process and improvements will be consequential. These metrics will be collected where possible from both completed trials and on-going trials. The expectation is that the automation provided by the MatchMiner platform will reduce the time to reach enrollment goals by ensuring no eligible patients are overlooked and minimizing the number of manual reviews required of the oncologists.

Written by