The Alliance for Clinical Trials in Oncology, one of the five cooperative research groups in the U.S. sponsored by the National Cancer Institute conducting large multicenter trials in oncology, and GNS Healthcare (GNS), a precision medicine company that applies causal machine learning technology to match health interventions to individual patients and discover new target intervention pathways, have announced a big data initiative.
The initiative is designed to identify patient subpopulations that may respond better to combination treatments for metastatic colorectal cancer. The collaboration leverages REFS (Reverse Engineering and Forward Simulation), the patented GNS machine learning and simulation platform.
REFS will analyze multi-modal genetic, genomic and clinical datasets associated with a recently completed phase III Alliance-led clinical trial of 1,137 patients that evaluated the efficacy of combination chemotherapies with biological agents. Specifically, the study evaluated Irinotecan/5-FU/Leucovorin or Oxaliplatin/5-FU/Leucovorin with Bevacizumab or Cetuximab (C225) in patients with advanced colorectal cancer. Irinotecan is sold by Pfizer under the brand name Camptosar and now is available as a generic. Oxaliplatin is sold under the brand name Eloxatin by Sanofi and now is available as a generic. Bevacizumab is sold by Genentech/Roche under the brand name Avastin. Bristol-Myers Squibb, Eli Lilly and Merck sell Erbitux, the brand name for Cetuximab.
“The complexity of this initiative—the robust datasets, as well as the level of scientific rigor needed to deliver results in a timely fashion—requires a partner with the most advanced computing and machine learning capabilities,” said Dr. Alan P. Venook, study chair for the Alliance, Madden Family Distinguished Professor of Medical Oncology and Translational Research at the University of California San Francisco, and Shorenstein associate director for program development at the Helen Diller Family Comprehensive Cancer Center. “The GNS machine learning platform has the ability to analyze extraordinarily large and complex datasets and to quickly deliver useful answers on the efficacy of combination treatments for metastatic colorectal cancer.”
The large-scale Alliance datasets include comprehensive clinical information and molecular profiles of patients diagnosed with metastatic colorectal cancer. The clinical parameters encompass demographics, patient and disease characteristics, and outcomes (e.g., overall survival). The molecular results represent detailed analyses of tumor tissue and blood specimens. All information and specimens were collected from patients who participated in the Alliance study from 2004 through 2014.
The effort hopes to accelerate and advance research on the efficacy and value of combination treatments for colorectal cancer with a goal to deliver a reliable framework to rapidly assess the impact of those treatments on individual patients. It aims to identify biomarkers of specific patient outcomes and to predict the likelihood of those outcomes at an individual patient level. That knowledge would improve the ability to match patients to the treatment regimen with the greatest value for them—and to do so sooner.
