News – MatchMiner

August 2026

OPTIONS-ACTIVATE launches

Dana-Farber has launched OPTIONS-ACTIVATE, a randomized study testing whether proactive AI trial suggestions help more patients reach clinical trials.

When MatchMiner-AI detects that a patient's cancer is progressing, their oncologist receives an email with a ranked list of possible trials. The study compares this with standard MatchMiner-AI access. Its main measure is the share of new systemic treatments that are clinical trials.

It builds on an earlier Dana-Farber randomized trial, which found that AI could reliably spot patients likely to need a change in treatment, but that notifications about genomically matched trials alone did not raise enrollment. OPTIONS-ACTIVATE instead sends a ranked list of trials from MatchMiner-AI, drawn from the patient's whole record.

OPTIONS-ACTIVATE is part of ACTIVATE (AI-driven Clinical-trial Trial-Information and Viability Assessment Tool for EHRs), a research program led by Kenneth Kehl, MD, MPH, and funded by the National Cancer Institute (R37 CA295653).

ClinicalTrials.gov NCT07232043 and Dana-Farber trial listing 25-413.

June 2026

MatchMiner-AI takes second place in the 2026 Hearst Health Prize

MatchMiner-AI took second place in the 2026 Hearst Health Prize, awarded at UCLA Health Data Day.

The recognition highlights its approach: matching patients to trials from the full medical record, with openly shared models trained on synthetic data.

February 2026

MatchMiner-AI goes live at Dana-Farber

Dana-Farber providers can now look up AI-suggested clinical trials for their patients, based on the full medical record.

MatchMiner-AI reads a patient's clinical notes and matches them against all open trials, not only those with genomic criteria. Each suggestion comes with a plain-language summary for the oncologist to review.

April 2025

JAMA Network Open publishes AI-triggered trial notification study

Our study, Clinical Trial Notifications Triggered by Artificial Intelligence–Detected Cancer Progression, is out in JAMA Network Open.

Patients often become eligible for a new trial when their cancer progresses. In this randomized trial at Dana-Farber, run from January 2023 to June 2024, AI applied to imaging reports detected likely progression, and the patient's oncologist received an email about genomically matched trials.

The AI reliably spotted patients likely to need a new treatment, but the notifications did not increase trial enrollment. That result shaped the design of the follow-up study, OPTIONS-ACTIVATE.

Watch Ken Kehl discuss the study (JAMA Network, on YouTube).

December 2024

MatchMiner-AI preprint posted on arXiv

MatchMiner-AI: An Open-Source Solution for Cancer Clinical Trial Matching is now available on arXiv.

MatchMiner-AI extends MatchMiner beyond genomic matching. It uses language models to read the clinical record and compare patients against trial eligibility criteria as written.

September 2024

MatchMiner-AI selected for a Llama Impact Grant

MatchMiner-AI was selected for the inaugural cohort of Meta's Llama Impact Grants. The grant supports development of an open-source AI platform for clinical trial matching.

The project uses Llama models to summarize unstructured clinical notes and trial eligibility criteria, so that suitable trials can be found for each patient. The models are integrated into MatchMiner and evaluated with clinical partners.

August 2022

MatchMiner paper published in npj Precision Oncology

MatchMiner: An open-source platform for cancer precision medicine describes how MatchMiner works and how clinicians and trial staff at Dana-Farber use it.

By January 2022, Dana-Farber had about 170 open precision medicine trials in MatchMiner, 84% of patients had at least one trial match, and more than 220 patients had enrolled on a trial as a result of a MatchMiner match.