Policy Day

BIO2003: Reporter's Notebook

Ronald Bailey | June 25, 2003

Washington, DC.—On Tuesday, the second full day of the world's largest biotech confab, I devoted myself to public policy and bioethics. Some small interesting bits of science and technology did, however, sneak in.

One side note—unlike previous gatherings, I have not seen a single anti-biotech protester at this year's conference. The organizers of BIO 2003 must be very pleased with themselves for scheduling their event at the same time as the Sacramento Ministerial, which is attracting a bit of anti-biotech protest activity, though less than the activists had hoped.

Who Owns the Genome?

This early morning panel was packed with legal talent discussing the mess that biotech patents are now in. Early on in the development of biotech patents, merely characterizing the genes and a method for isolating them was enough to claim patent protection. This led to the Gene Patent Rush of the 1990s, which made the California Gold Rush look tame. Now, it's coming time to sort out where scientists and corporations stand.

Discussing patent law, as valuable as that is, causes the eyes to glaze over faster than counting sheep. To cut a long session short, Barbara Caulfield, the general counsel for the gene-chip maker Affymetrix, warned that her company was hearing from researchers that patent licensing hassles were distorting their research agendas. "If a researcher wants to investigate proteins 1, 2, 3, and 4 for their roles in a disease," she said, "and there is an aggressive patent holder for protein 1, the researcher will skip 1 [even if it's more promising] and do 2, 3, and 4 first."

Affymetrix is advocating a controversial shift away from the current system, in which entities like individual genes and proteins can be patented, to one in which things like genes, proteins, haplotypes, single nucleotide polymorphisms (SNPs), and exons are treated as "catalog items"—no more patentable than elements like iron, carbon, and oxygen. Patents, Caulfield argued, should be awarded to functional inventions, such as new diagnostic tests or new drugs. Shifting away from genes to tangible inventions, she claimed, would benefit industry in the long run by providing "more predictability, more rationality, and more stability," which would in turn improve the biotech investment climate. Still, Caulfield acknowledges that some 70 percent of human genes are available, patent-free, in public databases.

Lawyer Reid Alder of the J. Craig Venter Science Foundation looked 10 years into the future, "when each of us have our whole genome available on CD ROM or USB chip or whatever storage media available then. We will then also be able to check the latest genomic research online for new studies to see if our genetic profiles reveal any particular disease susceptibilities." Adler believes that the future medicine involves finding disease susceptibilities based on genetic profiles of combinations of hundreds of genes. But such genetic profiling will only be possible if the testing companies are able to include a large number of probes for a wide variety of gene variants on the test biochips. Adler pointed out that, as of today, researchers have already developed 984 genetic tests—626 for diseases, and 358 for research purposes. If someone wanted to combine all these tests, they would have to seek licenses from the patent holders of all 984.

"What we don't want is for people to throw up their hands and not do genomic research because 984 people want 1 percent of the combined test's income stream," Adler said. "It can't work like The Producers, where every investor gets half of the income from a Broadway show." In this case, at 1 percent each, that would add up to 984 percent. "We need to encourage cost-effective mechanisms to consolidate patent rights that might impede genomic medicine," he declared.

Adler is also concerned because recent court decisions have apparently narrowed the "research exception," in which researchers could use patented products and processes without paying for a license. Another court decision, he said, is limiting researchers' "FDA exemption"—unlicensed use of patented products and processes for the purpose of creating data for submission to the Food and Drug Administration. He thinks both of these decisions will slow and distort research agendas.

Robert Blackburn, vice president and chief patent counsel for Chiron, advised patience. In his view, meddling now would only make things worse. Let the court decisions catch up to where the science is now, is his advice. Blackburn claimed that there is very little evidence that research is being prevented by the plethora of research patents. However, he offered a solution to any patent-created blockage that truly damages public health interests: "My analogy is the military jet," Blackburn said. "There are thousands of patents that cover a military jet, but there is no problem building one. If the U.S. infringes a patent, the only remedy is to go to the Court of Claims and ask for damages. Similarly, where health care is central, the National Institutes of Health should be able to infringe on a patent, and patent holders go the Court of Claims for damages."

To me, this "solution" seems a bit clunky and filled with uncertainties sure to scare off investors. Again, without wading into the briar patch that is biotech patent law, the Affymetrix proposal seems promising—simplistically put, it is a long-established principle that entities found in nature are not patentable. Things that are patentable are new, non-obvious, and useful. This provokes furious argument over exactly what those terms mean, but the fact is that for researchers it is now child's play to isolate new genes and characterize them—this, to me, seems to violate the non-obviousness provision of patentability.

Is There Genetic Discrimination?

Next up was a panel on the vexing question of what should be done about the specter of genetic discrimination. Robert Cook-Deegan from Duke University noted that the number of genetic markers is increasing enormously. When he was an Alzheimer's disease researcher in the 1970s, there were about 70 such markers, including things like the ability to roll one's tongue, and funny smelling urine after eating asparagus. As I heard earlier, there are now nearly 1,000 genetic tests.

The concern is that insurers and employers can use this genetic information to discriminate against employees and clients. Cook-Deegan did note that there is not much evidence to date that this is in fact happening. Later, the panelists would agree that there have been some weird outlier cases, e.g. when Burlington Northern had 36 employees tested for genes associated with carpal tunnel syndrome, with the apparent goal of denying them future compensation on the grounds that they had a pre-existing genetic condition. As we shall see at a later panel, genetic information is not peculiarly susceptible to misuse, and doesn't need any more protection than regular medical information.

Henry Desmarais from the Health Insurance Association of America agreed that there was little evidence that insurers are using genetic information to discriminate against clients. He also noted that the new Health Insurance Portability and Accountability Act regulations already protect the medical privacy of all Americans enrolled in group health plans. However, Desmarais did point to the problem of adverse selection. "Suppose there were a test for whether your house was going to burn down in the next 12 months," he offered. "Fire insurance wouldn't work well if the only people who bought it were people whose houses were going to burn down in the next year." Similarly, neither would health insurance if only people who knew that they were going to need it soon bought it. Insurance works because people are willing to take a small loss now in order to avoid the risk of a large loss later.