Biomarkers Consortium Launches Adaptive Trial To Test Targeted Breast Cancer Drugs

The five-year I-SPY 2 study is a model for drug development and an innovative screening process for identifying winners, FDA's Woodcock says.

The newly launched I-SPY 2 trial will put into practice many ideas about adaptive trial design and the use of biomarkers for patient selection, and it could herald the coming of a new model of drug development.

Under the sponsorship of the Foundation for the National Institutes of Health's public/private Biomarkers Consortium, the adaptive clinical trial will test at least a dozen oncology candidates from several companies in women with high-risk breast cancers and either advance or eliminate candidates after six months.

The trial, called "Serial studies to Predict Your therapeutic response with Imaging and molecular analysis 2" or I-SPY 2, launched this week at the University of California, San Francisco, one of about 20 cancer centers nationwide that will participate in the study, FNIH said during a March 17 press conference.

The five-year study is being conducted under a master IND held by FNIH. Under that umbrella, investigators will be able to graduate, drop and add drugs seamlessly throughout the trial without having to write a new protocol.

Graduated candidates will go back to their sponsors with a target population identified and a measure of certainty that a Phase III registrational study, with possibly as few as 300 patients enrolled, will be successful.

The ultimate goal is to speed breast cancer drug development, reduce failure costs and advance drugs targeted to work more effectively in individual patients. The study is expected to cost about $26 million, with approximately 60 percent of the funds already in hand, including undisclosed contributions from the Safeway Foundation, Johnson & Johnson, Genentech and Eli Lilly.

Biomarkers From Individual Patients' Tumors Are The Key

Initially, biomarkers from individual women's tumors will be used to screen drug candidates and identify a promising therapy. Each patient will then be given standard therapy plus that experimental agent for six months prior to surgery, while the tumor's response to treatment is measured periodically using MRI scans. (A reproducible method for using imaging to quantify the response to therapy was tested in I-SPY 1.)

The success or failure of the drug will be judged by how much of the tumor remains at the time of surgery. If the drug has at least an 85 percent chance of being successful in a standard, controlled Phase III trial "it graduates," Donald Berry, one of the two principal investigators on the trial explained. Berry, who chairs the department of biostatistics at the MD Anderson Cancer Center, designed the trial protocol after FNIH sought input from a broad stakeholder base.

As the trial progresses, investigators will be able to use data from each patient's treatment to predict whether a particular drug will work in other patients. Using this biomarker approach, "we'll learn from every patient, not wait for the end of the trial" to learn what works, Anna Barker, NCI deputy director and co-chair of the Biomarkers Steering Committee, said. Barker called the study "what the future of drug development should look like."

Approximately 800 women are expected to participate in the study altogether, with 120 patients at most in any one study arm. They will be followed for five years.

All results from the trial will be published by the investigators via articles in peer-reviewed scientific journals and via the I-SPY 2 study Web site. "Neither the funders or those donating drugs has any say over conduct or interpretation" of the trial, Charles Sanders, chairman of the FNIH board, stressed.

FNIH will serve as a trusted third party to manage the data and intellectual property that arise from the trial.

Studying Mechanisms, Not Drug Candidates

Five investigational agents from three drug makers have been qualified for the study so far, with many others lined up. To qualify, a candidate must have good safety data separately and in combination with chemotherapy. Another group of candidates will be announced early this summer, according to FNIH.

Each of the candidates represents a different drug class or mechanism type, so the trial is gathering data on drug classes, not specific drugs, Laura Esserman, UCSF, the study's other co-PI, explained.

"Instead of reinventing the wheel, we're actually taking this into a precompetitive phase and making sure that everyone has access to the information and no one has to repeat and pay for the studies again," Esserman said.

The first agents to be tested are ABT-888 (veliparib), a PARP inhibitor from Abbott; AMG 655 (conatumumab), an APO/TRAIL agonist, and AMG 386, an angiogenesis inhibitor, from Amgen; and CP-751,871 (figitumumab) an IGFR inhibitor, and HKI-272 (neratinib), a Pan ErbB inhibitor, from Pfizer. The sponsors are donating the drugs.

Not So Much A Trial As An Innovative Screening Process

FDA's Janet Woodcock called I-SPY 2 not so much a trial as an "innovative screening process" for evaluating the promise of breast cancer drugs.

Although it is definitely an important breast cancer trial, the study is "a model for innovation in drug development, something that is seriously needed," the Center for Drug Evaluation and Research director said.

"Too many discoveries languish because of the risk, cost, time and effort" required to evaluate a single candidate in the clinic, she said. "By correlating genetic signatures with tumor responses ... I-SPY will teach us how to effectively use biomarkers to enable targeted therapy."

By using adaptive trial design, I-SPY will enable "rapid adjustments in the strategy of the trial, based on real-time information, so it can be nimble and adjust, go toward the winner, so to speak, and move away from strategies that are not succeeding," Woodcock added.

-Shirley Haley ([email protected])