Computer platform helps clinicians match patients with cancer to precision medicine trials
Click Here to Manage Email Alerts
As precision medicine and tumor genomic testing play increasingly important roles in cancer treatment, the challenge of matching patients with appropriate cancer clinical trials has become more daunting.
To better manage and streamline this complex process, the Knowledge Systems Group at Dana-Farber Cancer Institute has launched MatchMiner, led by Ethan Cerami, PhD, and Michael Hassett, MD, MPH.
MatchMiner helps clinicians and researchers identify potential matches between patients and targeted therapy trials based on genomic testing of tumors. A study conducted by researchers at Dana-Farber Cancer Institute showed the platform helped facilitate enrollment of approximately 1 in 5 patients who have been genomically profiled and entered onto precision medicine trials. It has also played a role in expediting patient enrollment in such trials by more than 20%.
“MatchMiner solves two problems — the first is that trial-matching platforms are generally proprietary, so we created an open-source software,” Harry Klein, PhD, scientist and co-lead author of the study, told Healio. “The second is that there is a lot of patient data and clinical trial eligibility data in precision medicine trials, and we needed a way to connect the two so we could offer patients more options.”
Klein and study co-lead author Tali Mazor, PhD, scientist in the Knowledge Systems Group at Dana-Farber, spoke with Healio about the need to simplify and improve the process of matching patients with suitable clinical trials based on tumor genomics.
Healio: How does MatchMiner work?
Klein: There is an algorithmic matching engine that matches patient genomic data and clinical data to curated clinical trial eligibility data. Trial eligibility data is formatted in a human readable programming language clinical trial markup language, or CTML. Matching is based on Boolean logic, and we can encode very specific trial eligibility criteria to make an appropriate match.
On its surface is the user interface, which is very intuitive. The match engine runs at Dana-Farber nightly and matches are updated daily, and the clinician is able to see patient trial matches in the user interface. All of that background content, the match engine and the clinical trial markup language, is all under the hood, so to speak.
Healio: Can most clinicians easily run this program and identify appropriate trials for their patients?
Klein: There would need to be a bit of team infrastructure in place to get the data into the system, and then someone else would need to help curate trial eligibility.
Mazor: To follow up on that, once everything is set up, the clinician doesn’t have to do anything more than go to a website, or even the electronic health record (at Dana-Farber Cancer Institute, we integrated MatchMiner into the EHR). So, clinicians don’t even need to run the program. They just pull up a website and it’s all there. We’re doing all of that running stuff in the background. So, to Harry’s point, it’s very simple for the end user.
Healio: How did you evaluate MatchMiner in your study?
Klein: We did a retrospective analysis of a select group of consents who we called MatchMiner consents. When we compared the time to consent for those MatchMiner consents with other contents to the same trials, we found there was a reduction of approximately 55 days in time to consent. We also found MatchMiner facilitated about 20%, or 1 in 5 of the patients who had genomic profiling at Dana-Farber and enrolled onto genomically driven clinical trials. That is good compared with some of the lower accrual rates, typically 10% to 15%, in precision medicine trials.
Healio: What are the implications of this? Do you plan for it to be widely used?
Klein: We are certainly hoping it will be used widely. As far as our next steps are concerned, we are continuing to find improvements for the software itself and make those improvements known. We are also reaching out and talking with other institutions that have read our publication and are interested in MatchMiner and the possibility of forming partnerships.
Mazor: We are always looking for ways to keep improving. One of the difficulties of a system like this is that it’s taking in all patients who have undergone genetic profiling, but not all patients are looking for a trial at any given time. They might be on therapy, or they might not have any evidence of disease at the moment. So, we’re working on a collaboration with Kenneth Kehl, MD, MPH, who is a thoracic oncologist here, as well as a researcher in the division of population sciences. He has developed artificial intelligence models that predict which patients are likely to need a new therapy in the near future. This is based on the text that has been written to accompany an imaging report. We are working with Dr. Kehl to integrate these predictions into MatchMiner. We’re very excited about that.
References:
- Computer platform helps match patients with cancer to trials of targeted therapy (press release). Available at: www.dana-farber.org/newsroom/news-releases/2022/computer-platform-helps-match-patients-with-cancer-to-trials-of-targeted-therapy/. Published Oct. 12, 2022. Accessed Jan. 4, 2022.
- Klein H, et al. NPJ Precis Oncol. 2022;doi:10.1038/s41698-022-00312-5.
For more information:
Harry Klein, PhD can be reached at Dana Farber Cancer Institute, 450 Brookline Ave., CLSB 11007, Boston, MA 02215; email: hrklein@ds.dfci.harvard.edu.
Tali Mazor, PhD, can be reached at Dana Farber Cancer Institute, 450 Brookline Ave., CLSB 11007, Boston, MA 02215; email: tmazor@ds.dfci.harvard.edu.
Collapse
Read more about
Click Here to Manage Email Alerts