Goal of Personalized Medicine for Cancer Goes Mainstream

Four experts discussed progress, hopes and challenges related to personalized cancer treatment Tuesday at UCSF’s new Helen Diller Family Cancer Research Building on the Mission Bay campus. J. Craig Venter, PhD, a human genome pioneer; Brook Byers, a leading biotech venture capitalist; Frank McCormick, PhD, FRS, the director of the UCSF Helen Diller Comprehensive Cancer; and Susan Desmond-Hellmann, MD, MPH, the former head of product development at Genentech and UCSF chancellor-elect, participated in a panel discussion as part of the celebration of the opening of the building, dedicated entirely to cancer research. Watch the panel discussion. Among the challenges discussed: How can researchers obtain and analyze the massive amount of human DNA needed to develop the knowledge base that will support the development of a new, more potent, and better targeted drug armamentarium? What can be learned from the abnormalities within the tumor in comparison to what can be learned about the normal human genetic variations carried by the person in whom the tumor has arisen? What are the business, regulatory and reimbursement model for new treatments – and new diagnostics? Can the cost of bringing personalized medicine to the drug marketplace be reduced? It used to be that large pharmaceutical firms swung for the fences, trying to find a cancer drug to treat all cancers, or at least all patients with a particular cancer. Scientists and physician researchers at academic medical centers were among the first to advocate a more personalized approach.

From Blockbuster to Personalized Medicine

Cancer arises as a result of normal genes acquiring abnormalities, a discovery made three decades ago by Nobel-Prize-winning UCSF researchers Harold A. Varmus, and J. Michael Bishop, MD – the current UCSF chancellor. Now it’s clear that individual tumors differ from one to the next in significant ways when it comes to exactly which genes become abnormal and drive tumor formation, growth, survival and deadly migration to distant tissues. A treatment that works for one patient will not necessarily work for another in whom the tumor grows due to a different pattern of abnormal genes. In fact, current cancer drugs are effective in only a minority of patients. A shift in thinking has become mainstream, even at large pharmaceutical firms. Instead of pursuing blockbusters, more effort is being directed to figuring out how to bring better targeted drugs to market. These drugs are intended for fewer patients, but they are expected to be much more beneficial than previously developed drugs have been for patients with the targeted tumor profile. Some have doubted whether pharmaceutical companies would embrace this niche approach, given the cost of drug development. But panelists noted that growing knowledge of the genetic and biochemical properties of individual tumors may save drug development costs. Clinical researchers will become better at selecting patients for clinical trials who are most likely to benefit, and a greater understanding of genetic abnormalities and drug targets will lead to fewer misguided drug development projects, they said. Desmond-Hellmann, who will become UCSF Chancellor on August 3, was with Genentech in 1998 when the company launched Herceptin, a breast cancer drug widely regarded as the first targeted cancer therapy. As developed, the drug was intended for fewer than 30 percent of women with breast cancer – only those with tumors that made an abundance of the Her2 protein targeted by the drug. More than a dozen targeted cancer drugs now are available, and hundreds more are in the drug development pipeline. Desmond-Hellmann said that the promise of improved efficacy – even for a smaller percentage of cancer patients – can motivate drug development, and that careful patient selection for highly effective drugs can lead to smaller, less costly clinical trials. “What’s not to be missed is how small a clinical trial you can do with targeted or personalized therapy….The efficacy of the clinical trials process and the ability to show a big difference – which you really have to do today – is really enhanced.” Desmond-Hellmann had just returned from Orlando, where she attended the annual meeting American Society of Clinical Oncology. She noted that researchers reported positive results from a lung cancer clinical trial for a drug developed by pharmaceutical giant, Pfizer, targeted to an abnormality found in only about five percent of lung cancers. The idea that companies won’t support promising drugs that will benefit only a minority is clearly erroneous, she said.

New Day for Diagnostics

Byers, a partner at Kleiner Perkins Caufield & Byers, has been a venture capital investor, entrepreneur and company builder since 1972. Byers talked about combining ideas and information to transform diagnostics to change the practice of medicine. “Diagnostics speaks to about 80 percent of treatment decisions, but diagnostic tests are pretty crude right now,” he said. Byers emphasized the role better diagnostics can play in preventing under- or overtreatment of cancer, and in choosing the best treatment. Byers outlined the success of Oncotype DX, a diagnostic test on a chip that he helped shepherd to market. Oncotype DX is engineered to detect patterns of genetic abnormalities within tumors. These patterns are prognostic and can help guide treatment decisions. While the potential for new, more powerful, gene-based diagnostics now has been established, there are many wrinkles to work out, Byers said. How best to test, regulate and pay for the new diagnostics, which are much more expensive to develop than previous generations of diagnostic tests, remains an evolving issue. The US Food and Drug Administration “is really not ready yet for advanced diagnostics,” Byers said. The biotechnology industry is still working out standards for evaluating diagnostic tests. Diagnostic tests have heretofore been reimbursed by insurers at a low rate, even though the new diagnostics stand to save the health care system money, he noted.

Focus on Tumor Genomes, As Well as Human Genomes

A growing number of studies of human genetic variability – spurred by the initial success of the Human Genome Project – reveal clinically important differences in how each of us normally metabolizes different classes of drugs. But cancer researchers, despite all they have already learned about tumor genetics, need to learn more about the complex array of genetic abnormalities that can arise in tumor tissue. Venter noted a new goal for cutting through the complexity of tumor genetics. Even within the same tumor, cells differ. Many tumors are thought to be immortal due to rare cancer stem cells, some of which are beginning to be identified in various tissues. Venter pointed to technological advances that may soon enable investigators to spell out the genome of a single cancer stem cell or any other human cell. “We can even do the genome on a single bacterial cell now,” he said. “We can amplify a single molecule of DNA. So I think doing individual stem cells, cancer stem cells, will be critical to the future, and that’s hopefully only a few years away.” Desmond-Hellmann, a board-certified oncologist, cited the importance of laboratory researchers and clinical researchers working together to solve problems, including the challenge of obtaining adequate genetic information from very small tumor samples to conduct a lab analysis that can help guide treatment decisions. Large biopsies are rarely done any more. “It’s hard on the patients,” Desmond-Hellman said. Instead, fine-needle aspiration biopsy has become the pre-surgical norm for obtaining tissue. “We want to do genomic testing, but we have too little tissue.” Lab researchers and clinical researchers working together may find a way to obtain the needed information from small samples, she said. Byers, a driving force in the historical and ongoing development of the Bay Area Biotech industry, said UCSF was well-equipped to play a critical role in solving the challenges presented. “UCSF is a remarkable institution – I think maybe the best in the world for interdisciplinary work. That’s how these problems are going to be solved. It’s biology, biochemistry, molecular biology, sequencing, genetic medicine, medicinal chemistry, biostatistics, bioinformatics – all these things that come together and are integrated to give us these solutions.” McCormick expressed the hope of offering every UCSF cancer patient more promising treatment, based on genetic profiling and careful tracking and analysis of clinical outcomes. “One of our dreams for the future is to have all of our patients come onto clinical trials,” he said.

Related Links:

Fast-Changing Field of Medical Genetics Embraces Personalized Medicine
Science Café - UCSF