Accelerating Identification and Regulatory Approval of Investigational Cancer Drugs

Abstract

THE DEVELOPMENT OF NEW DRUGS IS BECOMING INCREASingly expensive—and oncology drugs, in particular, have a high clinical failure rate. The current return on capital investment in drug development by US public companies was recently reported as less than 0.3%. The low probability of success, coupled with rapidly accelerating expenses, means that drug development is increasingly the purview of only 2 organization types: a few large companies and myriad small, venture capital–funded start-up firms. At an estimated cost of $1.0 billion to $1.8 billion for developing a successful new drug, funding for such risky ventures, particularly for oncology drugs, may diminish. The high cost of oncology drug development is not only an issue of finance but also occurs because many cancers are heterogeneous. The inability (or lack of explicit effort) to identify and incorporate specific disease subtypes into trial design inhibits the development of more cost-effective drugs that target specific populations. The major losers in this inefficient approach are the patients who would benefit from new treatments. This dilemma necessitates new clinical trial designs that account for the heterogeneity and complexity of the specific disease at the outset and fully recognize that the problem is better solved through collaboration vs competition. Precompetitive collaborations will serve to advance these goals and enable a more efficient model of drug discovery. At a recent workshop entitled “The Role of Precompetitive Collaborations in Advancing Regulatory Science and Enabling Evidence-Based Review,” stakeholders representing various constituencies involved in new drug development discussed the benefits of precompetitive collaboration in creating pathways for regulatory approval for drugs that successfully demonstrate a significant improvement in surrogate end points in screening phase 2 trials. Precompetitive collaboration involves cooperation among traditionally competitive stakeholders, who work together on projects that advance mutual interest without providing a competitive advantage for any single organization. Such collaboration can allow institutions to pool resources and expertise for the multidisciplinary research necessary to accelerate drug development and allow more rapid sharing of successes and failures, furthering progress toward a shared goal of identifying classes of agents and the subtypes of diseases for which they are effective. Asanexample,theI-SPY2TRIAL(InvestigationofSerialStudies to Predict Your Therapeutic Response With Imaging and MolecularAnalysis)modelwasdevelopedasaprecompetitive collaborationamongmultiple academic,pharmaceutical, biotechnology,governmental, andadvocate stakeholders. I-SPY2 uses an adaptive design, modular trial process for the purpose of concurrently screening phase 2 agents in women with stage 2 and 3 breast cancer who are at increased risk for cancer recurrence and death despite standard adjuvant treatment. In this setting, pathologic complete response (pCR), measuring thecompletedisappearanceof tumor in response to treatment prior tosurgical excision,maypredict recurrence-free survival (RFS)—a current regulatory standard for Food and Drug Administration(FDA)approval.Thetrialevaluatesdrugs,byclass, in the context of standard and emerging biomarkers to determinewhetherthosedrugscanimprovethechanceofpCRcompared with standard therapy. The trial is powered to detect a doubling of the log odds of pCR within a biomarker signature. Drugs that are considered successful when they complete the trial are predicted to have an 85% likelihood of success in a confirmatory randomized neoadjuvant trial of 300 patients with tumor that have the drug’s newly identified biomarker signature. I-SPY2wasbasedonearlierworkinI-SPY1(CALGB150007 and 150012/ACRIN 6657), a collaboration of the Specialized Programs of Research Excellence (SPOREs) and the National Cancer InstituteCooperativeGroups.Prior to starting I-SPY2, the consortium worked for several years to refine the clinical approach and surrogates for RFS at 3 years. The group also developed an infrastructure for data sharing and the methods to miniaturize molecular assays and maximize the number of assays that could be performed on small amounts of tissue. The consortiumbaseditscriteria foreligibilityontheresultsofI-SPY 1, which shows that, in biologically high-risk palpable breast cancer, pCR differs by subset and is more predictive by subset than it isoverall, demonstrating that theextentofoutcomeadvantage conferred by pCR is specific to tumor biology.