No one in biotech – industry, academia, regulators, nonprofits or patients themselves – disputes that cancer drug development is at a critical juncture. A raft of young companies has joined early movers in the space and the pipeline is filled with promising drugs. But the explosion of research in diagnostics and biomarkers – in conjunction with the recognition that cancer is an amalgamation of diseases rather than a handful of broad indications – is calling for a more stratified approach to drug development.

First, the good news: Cancer drug approvals are up. FDA data show that 42 new molecular entities (NMEs) and nine biologics were approved for the active treatment of cancer between 2000 and 2012, including 15 NMEs and five biologics between 2010 and 2012. Over the past three years, the FDA's Office of Hematology and Oncology Products approved more drugs than any other therapeutic area overseen by the agency, according to spokeswoman Stephanie Yao.

And more therapeutic candidates for cancer are moving into the clinic. According to data from the Tufts Center for the Study of Drug Development (CSDD), published in its May/June 2013 CSDD Impact Report, the number of new cancer drugs that began clinical studies increased by 50 percent from the period between 1993 and 1998 to the 1999-2004 time frame. Although the average success rate for cancer drugs entering clinical studies between 1993 and 2004 was only 13 percent, the success rate doubled between the first and second halves of the study period, from 9.9 percent between 1993 and 1998 to 19.8 percent between 1999 and 2004. Those numbers suggest drug companies are making headway in improving the cancer drug development process.

"Oncology is one of the most active areas in drug development today and has been for some time," said Kenneth Getz, director of sponsored programs at the Tufts CSDD and chairman of the nonprofit Center for Information and Study on Clinical Research, or CISCRP. One of every four drugs presently in development targets cancer-related illness, "which is remarkable, and has dictated some of the new directions in innovation," he added.

That trend is borne out in data tracked by Cortellis Clinical Trials Intelligence (CTI), where the vast number of clinical trials for cancer drugs is a hopeful indicator. Approximately 10 percent of the universe of cancer trials tracked by Cortellis CTI over the past 12 months are phase III studies, and nearly half are phase II, as indicated in Fig. 1 below, suggesting that most cancer compounds that have moved into humans – or existing drugs under investigation in additional cancer indications – have advanced beyond initial safety and pharmacokinetic studies.

But the downside is that approximately six cancer candidates still fail during clinical trials for every one that makes it to commercial launch, according to Cortellis data. Even completion of a given trial is no guarantee that a drug will move forward. Of more than 40,000 trials of cancer therapeutics tracked over time by Cortellis Competitive Intelligence, nearly one-third were completed while only 10 percent were terminated or suspended. That relationship seems to confirm anecdotal evidence that many compounds are shelved by biopharmas after early studies show less than robust results in a targeted indication.

'WE HAVE TO START SHARING WHAT WE'RE LEARNING'

The amorphous cancer drug attrition rate is driving much of the effort to improve clinical trial design. In a white paper released in June at the American Society of Clinical Oncology meeting in Chicago, the National Patient Advocate Foundation (NPAF), in collaboration with the Cancer Innovation Coalition (CIC), commended the FDA for recent advances in cancer drug approvals but noted that "continued work on expedited regulatory pathways remains imperative." The paper added that "more needs to be done to improve access to innovation and increase the speed with which innovative therapies are brought to the market," noting that cancer drugs average nine years to move from discovery to approval compared to an average of just two years for HIV drugs.

"Given the importance of treating and eliminating cancer, and its prevalence in our society, steps should be taken, in partnership with industry stakeholders and government officials, to expedite these processes," according to the NPAF and CIC.

"Our nation and our researchers are beginning to appreciate that we have to start sharing what we're learning," Nancy Davenport-Ennis, founder and chairwoman of the NPAF, told BioWorld Insight. "We can't just shut down a clinical trial and say, 'Oh well. Seven years invested and we didn't learn anything here and don't have anything to bring to market.' We've got to put that out in the marketplace. There could be someone else who could take that experience and those data, apply them to a current clinical trial and end up with a very favorable result."

Richard Pazdur, director of the Office of Hematology and Oncology Products in the FDA's Center for Drug Evaluation and Research, declined to comment, and Yao said the agency doesn't "play a role in investment decision-making nor can we tell companies where to devote their resources."

Still, the FDA is beginning to acknowledge that designing clinical trials for smaller patient subsets with rarer tumor types – a consequence of the mapping of the human genome and the advent of personalized medicine – requires more data sharing and a certain amount of creativity.

"Some of these genes, some of these biomarkers that we're examining represent only 5 percent of a small population," said Ellen Sigal, chairwoman and founder of Friends of Cancer Research. "When you screen 100 patients and only 5 percent of them have the relevant gene, what happens to the other 95 percent? The knowledge of the science that we have now is very complex."

'IT'S A VERY COST-EFFICIENT PROCESS, AND EVERYONE GAINS'

Two years ago, individuals from industry, academia, government and patient advocacy organizations sat down at a Friends of Cancer Research annual conference to discuss the regulatory challenges associated with cancer drug development. Pazdur suggested participants discuss the development of a master protocol for lung cancer, according to Sigal.

"We were a little surprised," she admitted. "We usually don't work on clinical trials issues. But he thought lung cancer was a real opportunity, because he was seeing these one-off trials proposed that would not advance the field."

Sigal contacted Roy Herbst, professor of medicine and chief of medical oncology and associate director for translational research at the Yale Cancer Center, a veteran cancer researcher. The two assembled a core team, which returned to the FDA to ask if such an undertaking could be conducted as a registration trial – a framework that would encourage buy-in from patients and drugmakers. Pazdur again gave them a green light.

Those meetings were the genesis of Lung-MAP (SWOG S1400). The multidrug, biomarker-driven squamous cell lung cancer clinical trial is using state-of-the-art genomic profiling to match patients to multiple substudies testing investigational treatments targeting genomic mutations that drive the growth of their cancers. Patients are being recruited whether their tumors contain high levels of c-Met protein; the tumor DNA has FGFR gene amplification, mutation or fusion; the tumor DNA has PIK3CA gene mutation; the tumor DNA has CCND1, D2 or CDK4 gene amplification; or even whether their tumors have none of those changes.

Instead of undergoing multiple diagnostic tests to determine eligibility for a multiplicity of studies, enrollees are tested just once according to the master protocol and assigned to one of the five trial substudies, each testing a drug from a different industry partner.

The Lung-MAP protocol entails shared information and infrastructure, resulting in more opportunity for patients to be enrolled in a clinical trial and better access for researchers to relevant enrollees based on their genomic profiles. The entire process is designed to reduce the time and resources that drive the costs of cancer drug development.

"Working with the National Cancer Institute and with the community, we had to consider what this trial would look like," Sigal told BioWorld Insight. "How would it work? What drugs would we use? What biomarkers would we use? How would we conduct next-generation sequencing?"

Sigal credited Pazdur's support as the driver to launch Lung-MAP, which confronted the delicate dance of attracting stakeholders with otherwise competing needs.

"People don't do public-private partnerships for these trials unless there's something they can't really do themselves," Sigal maintained. "For the companies, it would be very difficult without using the cooperative group structure in a trial like this to recruit the rare tumor types they need. They'd have to screen hundreds and hundreds of patients before they could get the patients they wanted in these biomarker-driven trials, so that was very attractive."

Lung-MAP is designed as a randomized phase II/III registration trial, offering the potential benefit of moving directly to filings. Researchers were lured by the appeal of 400 participating centers, enabling them greater access to studies and the opportunity to have their questions answered more quickly than in limited settings. In addition, the federal government kicked in funding for the effort through the NCI.

"The biggest impact is on patients, because once they get screened there is an arm for every single patient," Sigal said. "It's a very cost-efficient process, and everyone gains."

Herbst, who testified earlier this month before the Subcommittee on Health of the House of Representatives Energy and Commerce Committee on the 21st Century Cures initiative, views Lung-MAP as a template for the future of cancer clinical trials. (See BioWorld Today, July 10, 2014.)

"We're all familiar with the technology that allows us to sequence the genome and to identify mutations, but we're frustrated, especially as clinicians, with our ability to match that information to a treatment to help a patient," he told BioWorld Insight. "Equally significant is the fact that if we don't do this work in the form of clinical trials, it's not going to help get drugs approved, which will ultimately help patients around the country and around the world."

'EXTRAORDINARY OPPORTUNITY TO GET PATIENTS THE RIGHT DRUG'

In a February 2010 draft guidance, the FDA defined adaptive trial design as that which "includes a prospectively planned opportunity for modification of one or more specified aspects of the study design and hypotheses based on analysis of data (usually interim data) from subjects in the study" at planned time points during the experiment.

Even before the FDA officially weighed in, clinical investigators were hailing the approach as providing "more data per dollar" by showing reason to kill likely failures early so that resources could be conserved for other pipeline candidates.

But investors, and even some scientists, remain wary of the approach, which has yet to go before the FDA as part of a new drug application. That discomfort with adaptive design played out with respect to another master protocol, the I-SPY 2 (Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging And moLecular Analysis 2) trial conducted by Puma Biotechnology Inc., of Los Angeles.

With the FDA's blessing, I-SPY 2 was designed to test whether, based on Bayesian predictive probability, a regimen containing neratinib, a pan-HER tyrosine kinase inhibitor, as neoadjuvant therapy in metastatic breast cancer likely would prove statistically superior to standard therapy in an equally randomized, 300-patient confirmatory trial. The company reported initial findings from I-SPY 2 in December 2013, and additional data were reported in April – to mixed reviews – during an oral presentation at the American Association for Cancer Research (AACR) meeting in San Diego. (See BioWorld Today, Dec. 6, 2013, and April 9, 2014.)

Wall Street's uneasiness, or downright confusion, over the I-SPY 2 data dropped Puma's shares (NYSE:PBYI) 18 percent on April 7 following the AACR presentation.

I-SPY 2 is the only one of 10 ongoing trials with neratinib to use the adaptive trial design. Another "plain vanilla" study, ExteNET, reported solid data last week on the use of neratinib for the extended adjuvant treatment of breast cancer, nearly tripling Puma's shares (NYSE:PBYI) in a single day and lending more credence to the I-SPY 2 experiment. (See BioWorld Today, July 24, 2014.)

No matter the outcome of Lung-MAP – still in its early days – or the role of I-SPY 2 in testing one of neratinib's many potential indications, sources agree that adaptive trial design is here to stay.

"Generally speaking, we've been dealing with the issue of inefficiency in clinical trials for many, many years," Sigal said. "The reality is that, now, we have an extraordinary opportunity to get patients the right drug that can work for them."

Editor's note: See next week's issue as BioWorld Insight examines the business, cultural and regulatory challenges to adaptive design in cancer trials.