Should the pharmaceutical industry continue its current path of
largely developing drugs that add only incremental increases in
efficacy, or more aggressively develop and tap knowledge of disease
processes to create medications that have a dramatic impact?
"I'm in favor of a clinical development approach where
hypotheses are derived directly from an understanding of the
pathogenesis of the disease, and a strong hypothesis leading into
clinical trials means a much lower risk of failure," said
William R. Sellers, MD, Global Head of the
Novartis Institutes for Biomedical Research, at a Special Session
of the American Association for Cancer Research (AACR) Annual
Meeting, held in Washington, DC.
Presentations at the Joint AACR/ASCO Forum made clear that 21st
century drug development will likely undergo major changes,
including in clinical trials.
Five Challenges
Dr. Sellers cited five major difficulties with targeting genetic
alterations in cancer:
- Most genetic alterations in cancer remain untreatable with
drugs, including genetic alterations in the ras and myc oncogenes.
"The pharmaceutical and academic communities will need to tackle
the problem of drugging these hard targets," Dr. Sellers said.
- A large number of genetic alterations occur in tumor suppressor
genes. There is a need to understand the interactions between
suppresser pathways and druggable nodes.
- Feedback loops can result in unanticipated signaling
activation.
- Cancers often have multiple pathways disrupted. Problems 3 and
4 need innovative approaches to identifying and developing novel
drug combinations, Dr. Sellers noted.
- Stratifying patients in clinical trials based on an
understanding of the target cancer's pathogenesis needs
improvement. "The trial logistics of that have been incredibly
hard, either from the perspective of getting biopsies or simply
executing the clinical trials in which you have to screen huge
numbers of patients to enroll a minor fraction," Dr. Sellers
said.
David P. Schenkein, MD, CEO of Agios
Pharmaceuticals, addressed the problems of failing to stratify
patients well.
"One consequence-in addition to perhaps keeping us in a mode of
incrementalism as opposed to making transformative effects with our
drugs-is whether we are actually throwing away important drugs,"
Dr. Schenkein said. He cited a simulation study that suggested that
without stratification in Herceptin [trastuzumab] trials, the
drug's benefits would have been missed. "Drugs currently failing
phase III studies and thrown away could potentially have been
transformative drugs if they had been subjected to the right
patient populations," he added.
Five Proposed Solutions
Dr. Schenkein offered five possible solutions:
- Gain a deeper understanding of the pathway interactions that
drive cancer tumors, because an effective drug needs to match a
cancer's genetic profile.
- Provide the same incentives and investments to the discovery of
diagnostics and biomarkers that are applied to discovering the drug
itself. Discovering the biomarkers at the time the drug is
discovered is essential.
- Change the clinical trial paradigm. Phase I trials need to
become more adaptive in design to test a variety of biomarkers and
perhaps imaging questions. That is happening more and more in
industry and academia, Dr. Schenkein said.
- Drug developers must work with the FDA to support adaptive
designs that will enable testing of biomarkers in phase I,
validation in phase II, and approval after a successful phase III
study. "Early clinical trials are going to have to be longer and
more expensive, and we all are going to have to agree to that," Dr.
Schenkein said.
- Industry and academia need to demand more important treatment
effects-which will dictate smaller treatment populations-and accept
nothing less.
Laurence H. Baker, DO, of the University of
Michigan, President of the Southwest Oncology Group, said an
internal review found SWOG needed to improve clinical trial designs
and operation efficiencies, and develop new partnerships.
This includes, for example, rethinking the definitions of
clinical benefit and progression-free survival (PFS) as endpoints.
"We clearly need to be more efficient," he added.
At the AACR meeting, Dr. Baker questioned the response
evaluation criteria in solid tumors (RECIST) definition of stable
disease (and, therefore, PFS). These criteria are "too permissive,
allowing companies to compare a new agent to a placebo and
demonstrating statistical difference but not clinical benefit," Dr.
Baker told The ASCO Post. For example, he noted, the many drugs
recently approved for metastatic renal cancer were evaluated using
such a design. "Now, some countries are not allowing use of some of
these agents because the costs are great and the benefit very
small. In other words, there is insufficient clinical benefit," he
said. ■
