A Look at I-SPY 2: Novel Trial Design May Expand the Scope of Oncology Drug Development

Oncology Live®, March 2014, Volume 15, Issue 3

Amid a growing recognition of the need to improve the process of developing oncology drugs, the novel I-SPY 2 clinical trial in breast cancer has demonstrated the potential to deliver new, effective treatment options more rapidly to patients who would most benefit while dramatically reducing the time and costs currently required to evaluate experimental therapies.

Amid a growing recognition of the need to improve the process of developing oncology drugs, the novel I-SPY 2 clinical trial in breast cancer has demonstrated the potential to deliver new, effective treatment options more rapidly to patients who would most benefit while dramatically reducing the time and costs currently required to evaluate experimental therapies.

The I-SPY 2 program employs an adaptive clinical trial design that allows for a new agent to be tested among smaller groups of patients identified as likely candidates for the therapy through detailed screening, and then uses statistical modeling to determine the probability of success in a larger study.

Need for Reforms Recognized

News of the first successful “graduations” from I-SPY 2 emerged as one of the major developments from the 2013 San Antonio Breast Cancer Symposium (SABCS) in December. Thus far, two compounds, the PARP inhibitor veliparib and the pan-ErbB inhibitor neratinib, are poised for phase III testing as a result of the program.Currently, clinical trials in oncology are designed to test just one drug at a time, and generally are focused on treating patients with advanced, metastatic disease, with success in metastatic disease followed by investigations in the adjuvant setting. The process of getting a drug to market can take a decade or more, hundreds of millions of dollars, thousands of patients—and still face a high likelihood of failure.

Only 1 in 10 drugs moves successfully from a phase I clinical trial to FDA approval, and oncology agents have the lowest likelihood of approval, according to a recent industry study of approval trends from 2011-2013. In fact, only 45% of phase III trials of oncology drugs result in FDA approvals.

Although there are numerous reasons for these failures, there is a growing suspicion among those involved in drug development that good drugs are being lost to inefficient clinical trial designs that are unable to appropriately test novel anticancer agents. It has been increasingly recognized that the neoadjuvant setting has significant potential to provide a unique platform to test new treatments much more rapidly. It offers a powerful opportunity to examine the effects of an experimental agent on a tumor before it is surgically removed, to codevelop biomarkers, and to individualize therapy.

However, there are significant safety and ethical considerations associated with testing investigational agents, with limited safety data, in the neoadjuvant setting in patients who are potentially curable. In the current model of drug development, the only way that these issues can be addressed is by ensuring that the drug has already undergone significant phase II studies in patients with advanced-stage disease. This is a time-consuming and expensive process that limits the potential for accelerated drug development.

I-SPY 1 Establishes Concept

The I-SPY trials represent a significant example of efforts under way to address the need for fundamental change in the clinical trial system. The program is sponsored by the Biomarkers Consortium, a public-private biomedical research partnership managed by the Foundation for the National Institutes for Health (FNIH), and involves a collaboration among stakeholders from industry, academia, government, and others. These are phase II trials in the neoadjuvant setting in patients with locally advanced, stage II/III breast cancer, designed to rapidly codevelop novel agents and biomarker signatures.

The I-SPY 1 trial was an initial study of neoadjuvant chemotherapy (NAC; doxorubicin/cyclophosphamide followed by paclitaxel) in women with high-risk breast cancer, with a principal objective of identifying indicators of response to NAC in this patient population.

I-SPY 2 At a Glance

Criteria for drug inclusion

  • Evidence of potential efficacy from preclinical/clinical studies
  • Tested and found safe in at least 1 phase I study with a taxane (or a taxane in combination with trastuzumab in patients with HER2-positive tumors)
  • Only 1 representative drug from a particular class

Criteria for “graduation”

  • Meet a threshold of 85% for predicted likelihood of success in a phase III trial of at least 300 patients
  • Double the log odds of achieving pathologic complete response

Active Clinical Investigations

Agent/Regimen

Target (s)

Company

Trebananib (AMG 386) ± trastuzumab (Herceptin)

Angiopoietin 1/2-neutralizing peptibody + HER2-targeting agent

Amgen

Ganitumab (AMG 479) + metformin

Insulin-like growth factor 1 receptor inhibitor + anti hyperglycemic agent

Amgen

MK-2206 ± trastuzumab

Akt inhibitor + HER2-targeting agent

Merck

Pertuzumab (Perjeta) + trastuzumab

HER2-targeting agents

Genentech

Pertuzumab + T-DM1 (Kadcyla)

HER2-targeting agents

Genentech

Graduated Agents

Veliparib (ABT-888) + carboplatin

PARP inhibitor with platinum compound

AbbVie

Neratinib

Pan-ErbB inhibitor

Puma Biotechnology

The study determined the role of a range of biomarkers on therapeutic outcomes and evaluated the longitudinal aspect of MRI imaging, according to Donald A. Berry, PhD, a professor in the Department of Biostatistics at The University of Texas MD Anderson Cancer Center who helped develop the design.

“Both sets of information were critical in I-SPY 2,” Berry said in an interview with OncologyLive. “Biomarker information was used to set the signatures in I-SPY 2. An MRI longitudinal model informed by results of I-SPY 1 was used to enable earlier decisions about graduation, futility, and randomization proportions in I-SPY 2.”

Furthermore, Berry explained, “Although it was not an explicit goal of the trial, it had a proof-ofconcept aspect. Namely, we showed that we could actually accrue early-stage breast cancer patients to a neoadjuvant trial in a multicenter setting. The neoadjuvant approach was still new to many of our investigators.”

I-SPY 1 is no longer accruing patients, but participants continue to be followed for long-term outcomes.

“It continues to be a rich source of information about biomarkers and imaging in neoadjuvant breast cancer and we continue to publish papers on the results,” Berry said.

I-SPY 2 Takes Next Step

The I-SPY 2 trial is designed to test multiple investigational targeted breast cancer therapies at the same time to determine whether adding each of these agents to standard NAC produces better outcomes than standard NAC alone. The trial involves a control arm that tests a standard NAC regimen plus trastuzumab for patients with HER2-positive tumors and multiple treatment arms for the investigative agents being tested (Figure). Berry noted that the order of NAC was reversed in I-SPY 2 in order to randomize experimental therapy to the paclitaxel cycles as early as possible.

Core biopsies, MRI scans, and blood and tissue sampling are performed throughout the study.

Core biopsies are used to determine biomarker categories, while MRI scans are used to predict pathologic complete response (pCR) rates. Patients are followed for up to 10 years to determine disease- free survival rates. The primary endpoint is pCR, defined as no residual disease in breast or lymph nodes at surgery.

Principal among the unique attributes of I-SPY 2 is its adaptive design, which allows investigators to learn from study data as they are collected and adapt accordingly. In 2010, the FDA recognized the burgeoning interest in adaptive trial designs by releasing a draft guidance to help promote and guide their use in clinical testing of investigational agents.

More specifically, I-SPY 2 has a multiarm Bayesian adaptive randomization design developed by Berry and Laura Esserman, MD, MBA, director of the Carol Franc Buck Breast Cancer Center at the UCSF Helen Diller Family Comprehensive Cancer Center.

Rather than waiting until the end of the trial, outcomes are assessed continually and accumulating data are used to inform randomization probabilities. Investigational agents “graduate” to smaller, confirmatory phase III trials if they reach a threshold predictive probability of superiority to standard NAC in a phase III trial, or are dropped from development if they have a low predictive probability.

Biomarkers Play Vital Role

Another novel aspect to the trial is its approach to addressing biomarker development. Biomarkers can be used to inform all aspects of drug development and have the potential to lower drug failure rates by tailoring treatment to patient populations that are more likely to benefit. However, the clinical validation of biomarkers has proved to be a significant challenge.

In the I-SPY 2 trial, investigational agents are codeveloped with three tiers of biomarker signatures: standard signatures, which are accepted and approved by the FDA; qualifying signatures, which are not yet FDA approved, but have substantial clinical data; and exploratory signatures, which are in the earliest stages of clinical validation.

Standard biomarkers are used to define the 10 biomarker signatures that are candidates for pairing with investigational agents for codevelopment, while data from the qualifying and exploratory classes will be evaluated retrospectively to potentially improve a drug’s biomarker signature.

Drugs graduate along with the biomarker signature in which they demonstrate success and, therefore, subsequent phase III trials can be much smaller and more precise.

In I-SPY 2, drugs are being evaluated against biomarker signatures consisting of combinations of hormone receptor (HR) status, HER2 status, and two levels of the MammaPrint 70-gene signature score. Although this produces 256 possible signatures, 10 have been chosen for the I-SPY 2 trial based on the biology they represent and their expected high prevalence in the study population.

Although many of these signatures have been recognized as potentially predictive of response to therapy in other studies and are standard signatures for the most part, Berry said that there are two unique aspects of the biomarker analysis in the I-SPY trials. “One is that they were recognized as being candidates for possible pairings with experimental agents, in moving toward personalized medicine,” he said. “The other is that these possible matches with the various experimental therapies are considered continually through the trial with graduating therapies in one or more of the 10 signatures occurring at any time.”

First Two Graduations

Recent reports of the first graduations from the I-SPY 2 trial have reinforced the potential utility of this trial design as a model for accelerated drug development. At SABCS, Hope S. Rugo, MD, director of Breast Oncology and Clinical Trials Education at UCSF Helen Diller Family Comprehensive Cancer Center, reported that the combination of veliparib and carboplatin plus standard therapy demonstrated a 90% predictive probability of delivering better outcomes for patients with triple-negative breast cancer compared with standard therapy in a 300-patient phase III trial.

After 6 months of treatment, 71 evaluable patients who received the veliparib combination demonstrated a 52% pCR versus a 26% pCR among the 62 patients in the control arm, which corresponded to a 99% probability of superiority to the control arm. There was a moderate increase in toxicity that was well managed.

In contrast, the pCR benefits were much smaller for the veliparib arm compared with the control group for all HER2-negative patients (33% vs 22%) and for HER2-negative/HR-positive patients (14% vs 19%).

One question that has arisen about the I-SPY 2 trial involves whether the benefit seen with the veliparib regimen can be fully attributed to the novel agent. The results of the CALGB/Alliance 40603 study, also reported at SABCS, indicated that the addition of carboplatin to standard chemotherapy demonstrated a similar pCR rate.

Berry said the I-SPY 2 trial used a more stringent definition of pCR than the CALGB study concerning surgery and nonprotocol therapy.

That aside, however, he believes the best way to compare the studies is via the difference between pCR rates between experimental and standard therapies. Those data indicate that the veliparib regimen had a comparatively greater impact on pCR rates, Berry said.

The second acknowledged graduate of the I-Spy 2 trial is neratinib, an irreversible tyrosine kinase inhibitor that blocks epidermal growth factor signaling. Neratinib was evaluated in combination with paclitaxel followed by AC. The trial included 115 patients, including 65 participants with HER2-positive disease.

The neratinib regimen demonstrated a 94.7% likelihood of superiority over standard neoadjuvant chemotherapy, and a 72.5% probability of success in a 300-patient phase III trial in comparison with this standard regimen for patients whose tumors were HER2-positive and HR-negative, according to Puma Biotechnology, Inc, which is developing the drug.

Ethical Issues Addressed

As I-SPY 2 unfolds, researchers have been mindful of the ethical considerations of testing investigational agents in patients who have other treatment options.

The methods used for selecting which investigational agents will be evaluated have helped to address the safety issues related to testing these agents in the neoadjuvant setting. An independent expert committee selects the drugs to be evaluated. The trial has also benefited from the involvement of breast cancer advocates, who have helped to create brochures, a website, and a DVD to ensure that patient consent is well informed.

In 2012, the President’s Council of Advisors on Science and Technology singled out the I-SPY trials as pioneering approaches to improving the efficacy and speed of medical research. That same year, the FDA developed a new guidance on the use of neoadjuvant trials as a pathway to accelerated drug approval.

Under this new guidance, I-SPY 2 results, in combination with phase III follow-up studies, could lead to accelerated approval of the investigational agents being studied. Meanwhile, the I-SPY2 leaders are also developing guidelines of their own, to inform other researchers how to perform trials like this and to mitigate the risks of testing experimental drugs in a potentially curable population.

Figure. How I-SPY 2 Tests Experimental Regimens

AC indicates doxorubicin (Adriamycin) and cyclophosphamide; ER, estrogen receptor; HER2, human epidermal growth factor receptor 2; HR, hormone receptor; PR, progesterone receptor.

Key Research

Alexander BM, Wen PY, Trippa L, et al. Biomarker-based adaptive trials for patients with glioblastoma—lessons from I-SPY2. Neuro Oncol. 2013:15(8):972-978.

Barker AD, Sigman CC, Kelloff GJ, et al. I-SPY 2: an adaptive breast cancer trial design in the setting of neoadjuvant chemotherapy [published online May 13, 2009]. Clin Pharmacol Ther. 2009;86(1):97-100.

Berry DA, Herbst RS, Rubin EH. Reports from the 2010 Clinical and Translational Cancer Research Think Tank meeting: design strategies for personalized therapy trials. Clin Cancer Res. 2012;18(3):638-644.

DeMichele A, Berry DA, Zujewski J, et al. Developing safety criteria for introducing new agents into neoadjuvant trials [published online March 7, 2013]. Clin Cancer Res. 2013;19(11):2817-2823.

Hay M, Thomas DW, Craighead JL, Cella Economides, Rosenthal J. Clinical development success rates for investigational drugs [published online January 9, 2014]. Nat Biotechnol. 2014;32(1):40-51.

Ledford H. Four ways to fix the clinical trial. Nature. 2011;477(7366):526-528.

Printz C. I-SPY2 may change how clinical trials are conducted: researchers aim to accelerate approvals of cancer drugs. Cancer. 2013;119(11):1925-1927.

Puma Biotechnology reports positive top line data from I-SPY 2 trial [news release]. Los Angeles, CA: Puma Biotechnology, Inc; December 4, 2013. http://www.pumabiotechnology. com/pr20131204.html. Accessed February 19, 2014.

Rugo HS, Olopade O, DeMichele A, et al. Veliparib/carboplatin plus standard neoadjuvant therapy for high-risk breast cancer: first efficacy results from the I-SPY2 Trial. Presented at: 2013 Annual San Antonio Breast Cancer Symposium. December 10-14, 2013; San Antonio, TX. Abstract S5-02.