VANCOUVER, British Columbia – Progress in cancer drug development depends on patients who have exhausted existing therapies agreeing to take part in clinical trials of new drugs.

Translating the potential of cancer genomics is a particularly stirring example of that selflessness in action, as Daniel von Hoff, professor of medicine at the Mayo Clinic and also of the Translational Genomic Research Institute of Scottsdale, Ariz., said at the American Association for the Advancement of Science meeting over the weekend.

"All the patients we see have exhausted all standard therapies and have a life expectancy of around 12 weeks," von Hoff told delegates. The tumors of each patient are subjected to molecular profiling to assess if they are expressing any genes for which targeted treatments exist, or are in clinical development.

Von Hoff displayed a list of the many patients who died in 2011. But, he said, "there have been some triumphs too, and these are becoming more common." A number of patients have had complete responses after the profile of their tumors highlighted mutations against which there is a targeted drug.

Most of those responses have been in rare cancers, though there are examples of more common cancers responding, too. One example was finding a young woman, a nonsmoker who had a lung cancer expressing EGFR (epidermal growth factor receptor). "It is becoming clearer and clearer that you need to look for the Achilles heel or a 'context of vulnerability'," von Hoff said.

If profiling the tumors of those patients is to be of any value, there have to be compounds with mechanisms of action that are relevant to the mutations that are uncovered. Von Hoff told delegates the Mayo Clinic carries out a huge number of first-in-human trials of cancer drugs in development, including testing such biotech industry favorites as PARP inhibitors, aurorakinases, topoisomerases and hedgehog pathway blockers.

In a formal clinical trial, in which the tumors of 112 patients with end-stage disease were profiled, 74 percent exhibited targets against which there are approved drugs. Following that, von Hoff is leading a prospective study, being conducted at nine centers in the U.S., comparing progression-free survival in patients given conventional therapy vs. patients treated according to the targets thrown up by molecular profiling. Alarmingly perhaps, there was little correlation between the treatment options highlighted by profiling and the treatments that clinicians selected.

Von Hoff said he believes that is a "good start" in applying understanding of the genetics of tumors to select the appropriate treatment. "No one is naïve enough to think that what we have done so far is a big advance; it's an early approach, but it did help some patients."

More progress is possible because of scientific advances that are uncovering new mutations and pathways, and leading to the development of new inhibitors. Several of the major U.S. cancer institutes routinely sequence the tumors of every patient and von Hoff is now moving on from molecular profiling to doing whole genome sequencing and transcriptome analysis of refractory tumors. "This is only the beginning, in terms of targeted treatment," von Hoff said, noting that there already are some benefits of complete genome sequencing, including avoiding the need to take repeated biopsies.

"To get the best breakthroughs, we need the best possible science," von Hoff concluded. "If we are profiling patients' cancer, we need to have relevant compounds."