Deep Clinical Trajectory Modeling to Optimize Accrual to Cancer Clinical Trials

This study aims to evaluate the effectiveness of proactive notifications to treating oncologist to optimize participant accrual to clinical trials by utilizing the MatchMiner AI platform. This study compares the standard MatchMinder AI access method to two enhanced recruitment methods.

Study Overview

• Status: Completed
• Conditions: Cancer
• Intervention / Treatment: Other: AI-assisted MatchMiner Platform

Detailed Description

The goal of this medical record data analysis and health system implementation study is to evaluate the effectiveness of proactive notifications to treating oncologist to optimize participant accrual to clinical trials by utilizing the MatchMiner platform. This study compares the standard MatchMinder access method to two enhanced recruitment methods.

In the first phase, investigators will provide qualitative feedback to improve AI algorithm impact on clinical trial accrual and the delivery of information from the MatchMiner platform that is utilized by treating oncologists and investigators.

In the second phase, medical records identified by the MatchMiner platform as available or a "match" for clinical trial enrollment will be randomized into three cohorts with the randomization occurring at the participant level. In Group 1, treating oncologists can use MatchMiner in its traditional form to identify potential clinical trial candidates based on structured genomic data and cancer type. In Group 2, treating oncologists will automatically receive emails with lists of potential genomically matched clinical trials identified by MarchMiner for patients in whom our AI algorithm detects an elevated probability of changing treatment based on imaging reports; oncologists can also still use traditional MatchMiner workflows. In Group 3, treating oncologists will receive email lists of genomically matched clinical trials identified by Matchminer for patients with AI-detected elevated probability of treatment change, after additional manual review to confirm that patients had progressive diseased based on their imaging reports and did not meet one of the common exclusion criteria for most cancer trials (including uncontrolled brain metastases, multiple primary cancers, poor performance status, lack of measurable disease, already having changed treatment, and hospice enrollment).

Of note, this study was not itself considered a clinical trial during the initial NCI grant application process or on subsequent discussion with NIH staff, since the outcomes were research processes (whether patients enrolled in other therapeutic clinical trials), not health-related patient outcomes as per the NIH definition of a clinical trial. However, for publication, a medical journal determined that the study met ICMJE criteria for a clinical trial and requested that it be registered.

• Study Type: Interventional
• Enrollment (Actual): 20707
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • Massachusetts
    • Boston, Massachusetts, United States, 02215
      • Dana-Farber Cancer Institute

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

-≥ 18 years of age

-adults with any type of cancer whose tumors underwent OncoPanel genomic sequencing from 2013-2022

Exclusion Criteria:

-≤ 18 years of age.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Other
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

3

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
No Intervention: Group 1: MatchMiner; Treating oncologists and investigators can use the standard method of accessing the MatchMiner tool to identify potential clinical trials for eligible participants based on structured genomic criteria.
Experimental: Group 2: MatchMiner Proactive Notification based on AI-detected progression; treating oncologists will automatically receive emails with lists of potential genomically matched clinical trials identified by MarchMiner for patients in whom our AI algorithm detects an elevated probability of changing treatment based on imaging reports; oncologists can also still use traditional MatchMiner workflows.	Other: AI-assisted MatchMiner Platform; A medical record data analysis tool that uses conjunction machine learning and natural language processing models to predict changes in treatment and prognosis and ascertain progression of disease and metastatic sites using retrospective imaging reports. MatchMiner is an established clinical operations tool at Dana-Farber Cancer Institute that links OncoPanel next-generation sequencing data to basic clinical information and clinical trial eligibility criteria to suggest biomarker-selected therapeutic trials for participants.
Experimental: MatchMiner AI with Proactive Notification Based on AI-detected progression + Study Team Confirmation; treating oncologists will receive email lists of genomically matched clinical trials identified by MatchMiner for patients with AI-detected elevated probability of treatment change, after additional manual review to confirm that patients had progressive diseased based on their imaging reports and did not meet one of the common exclusion criteria for most cancer trials (including uncontrolled brain metastases, multiple primary cancers, poor performance status, lack of measurable disease, already having changed treatment, and hospice enrollment)	Other: AI-assisted MatchMiner Platform; A medical record data analysis tool that uses conjunction machine learning and natural language processing models to predict changes in treatment and prognosis and ascertain progression of disease and metastatic sites using retrospective imaging reports. MatchMiner is an established clinical operations tool at Dana-Farber Cancer Institute that links OncoPanel next-generation sequencing data to basic clinical information and clinical trial eligibility criteria to suggest biomarker-selected therapeutic trials for participants.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Percentage of Patients Enrolling in Any Dana-Farber Cancer Institute Therapeutic Clinical Trial of Anti-Cancer Systemic Therapy	This measure assesses the proportion of patients in each study arm who enroll in any Dana-Farber Cancer Institute (DFCI) therapeutic clinical trial involving anti-cancer systemic therapy during the intervention period. Trial enrollment data will be pulled from the institutional OnCore database.	Up to 18 months

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Number of Patients Having Consultations with the Center for Cancer Therapeutic Innovation (CCTI)	Counts the number of patients in each study arm who have at least one consultation with the CCTI at Dana-Farber Cancer Institute during the intervention period. Encounters with the CCTI will be pulled from the institutional Enterprise Data Warehouse.	Up to 18 months
Percentage of Patients Consenting to Any Clinical Trial of an Anti-Cancer Systemic Therpay	Counts the proportion of patients in each study arm who provide consent to participate in any therapeutic clinical trial during the intervention period. Trial consent data will be pulled from the institutional OnCore database.	Up to 18 months
Percentage of Patients Predicted to Change Treatment Who Enroll in Any Therapeutic Clinical Trial	Assesses the proportion of patients identified by our AI model as likely to change treatment within the next 30 days who subsequently enroll in any DFCI therapeutic clinical trial involving anti-cancer systemic therapy. Trial enrollment data will be pulled from the institutional OnCore database.	Up to 18 months
Percentage of New Systemic Therapy Initiations That Are Clinical Trials of Anti-Cancer Systemic Therapies	Determines the proportion of new systemic therapy regimens initiated during the intervention period that are part of a therapeutic clinical trial. New systemic therapy initiations will be pulled from the institutional Enterprise Data Warehouse, and clinical trial enrollment data will be pulled from the institutional OnCore database.	Up to 18 months
Clinician Opt-Out Rate from Ongoing Email Notifications	Measures the percentage of clinicians in each intervention arm (Groups 2 and 3) who opt out of receiving ongoing email notifications from the study. This will be measured using physician survey responses.	Up to 18 months
Comparison of Anti-Cancer Systemic Therapy Clinical Trial Enrollment Proportions Between AI-Assisted Intervention Arms (Groups 2 and 3)	Compares the percentage of patients enrolling in any DFCI therapeutic clinical trial of anti-cancer systemic therapy between Group 2 and Group 3 to evaluate the effectiveness of automated notifications versus notifications after manual review. Trial enrollment data will be pulled from the institutional OnCore database.	Up to 18 months

Collaborators and Investigators

• Sponsor: Dana-Farber Cancer Institute
• Collaborators: National Cancer Institute (NCI)
• Investigators: Principal Investigator: Kenneth Kehl, MD, MPH, Dana-Farber Cancer Institute

Study record dates

Study Major Dates

• Study Start (Actual): January 30, 2023
• Primary Completion (Actual): July 15, 2024
• Study Completion (Actual): July 15, 2024

Study Registration Dates

• First Submitted: March 11, 2025
• First Submitted That Met QC Criteria: March 14, 2025
• First Posted (Actual): March 25, 2025

Study Record Updates

• Last Update Posted (Actual): March 25, 2025
• Last Update Submitted That Met QC Criteria: March 14, 2025
• Last Verified: March 1, 2025

More Information

Terms related to this study

• Keywords: Cancer
• Other Study ID Numbers: 19-536; 1K99CA245899 (U.S. NIH Grant/Contract); R00CA245899 (U.S. NIH Grant/Contract)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: The Dana-Farber / Harvard Cancer Center encourages and supports the responsible and ethical sharing of data from clinical trials. De-identified participant data from the final research dataset used in the published manuscript may only be shared under the terms of a Data Use Agreement. Requests may be directed to: Dr. Kehl. The protocol and statistical analysis plan will be made available on Clinicaltrials.gov only as required by federal regulation or as a condition of awards and agreements supporting the research.
• IPD Sharing Time Frame: Data can be shared no earlier than 1 year following the date of publication
• IPD Sharing Access Criteria: Contact the Belfer Office for Dana-Farber Innovations (BODFI) at innovation@dfci.harvard.edu
• IPD Sharing Supporting Information Type: STUDY_PROTOCOL; SAP

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No