Appeals Court overrules lower court, upholds breast cancer gene test

By John Timmer

A bit over two years ago, a large collection of interested parties joined together with the ACLU to challenge a practice that had become business as usual: the patenting of human genes. The group filed a suit that targeted a specific set of patents: those used by Myriad Genetics to control the market for tests of the BRCA breast cancer genes. In a surprise result, the judge overseeing the case issued a sweeping ruling that not only voided Myriad's patents, but seemed to put all gene patents at risk. Now that decision has been reversed, as an Appeals Court has upheld the patents in question.

The initial ruling that voided Myriad's patents relied on a novel interpretation of what was being patented. Patents on natural substances are allowed if the process of obtaining them is transformative, meaning the end result differs significantly from the original, natural state. In order to do genetic testing, the judge reasoned, it didn't matter which particular DNA molecule was obtained or in what manner—what you needed was the information conveyed by the gene, in terms of its protein sequence, not a specific piece of DNA. And that information is a natural product, which can't be patented unless it was transformed in some way. Since the genetic test doesn't involve any transformations, then the patents were invalid. Since many other gene-focused patents relied on this sort of information, rather than the actual DNA, this decision seemed to place many of them at risk.

We noted this was a rather novel interpretation of gene patents, and it clearly went against the initial decision on the topic by the US Patent and Trademark Office, as well as decades of practice since that decision. But to a certain limited extent, it has been adopted by the Appeals Court, which did throw out some aspects of Myriad's patents.

The aspects of the patents that didn't survive this latest round of scrutiny focus on what happens once you obtain a sequence from a test subject. At that point, to determine whether there might be any significant mutations, it's standard procedure to compare this sequence to one from a person who does not have mutations in the BRCA genes. The results of this comparison an essential part of the genetic testing procedure, and the comparison was included in some of Myriad's patents. The Appeals Court, however, accpeted the earlier ruling, which concluded the comparison was a information-focused, mental process, and thus wasn't eligible for patenting.

That, however, was the only part of the patents that didn't survive. Myriad's patents covered a way to test for chemicals that could be effective at slowing the growth of cells that carry BRCA mutations. Again, this was a standard laboratory procedure: grow cells in a culture dish, add the compound, and determine whether it influenced the trajectory of growth. Even though this is a very common approach to looking for cancer therapies, the court ruled it was transformative, and thus patentable. Although this only applies to cells carrying BRCA mutations, the ruling raises the risk that companies could patent the same procedure for any of the hundreds of genes associated with various cancers, so it wouldn't be surprising if something similar ended up back in court in the future.
Are genes patentable?

But the key part of the ruling focused on the patentability of genes. Here, the court focused on the process of isolating the DNA sequences of the genes. Normally, the genes reside on the cell's chromosomes, which are too clumsy to work with. So, instead, the patents focused on various methods of isolating smaller pieces of the genes that are easier to work with. As a result of these methods, the court ruled, the DNA sequences are chemically transformed, as they are no longer part of the long DNA polymer that makes up the chromosome. This process, it concluded, is sufficiently transformative to make it patent worthy.

The end result of this decision is that the patents stand, which allows Myriad to continue its monopoly on the genetic tests. The ruling, however, contained some awkward reasoning, a stinging dissent, and a half-hearted concurrence, all of which suggest that there may be significant wiggle room for attacking the patents by a different legal route.

The ruling focuses on how having a different arrangement of bonds in the DNA that is isolated is enough to distinguish it in its natural state. But the court was faced with briefs that suggested this was a dangerous line of reasoning, since elements like lithium are reactive enough that they only exist naturally as part of a chemical compound. Does this mean that someone can patent pure lithium? The court indicated the answer is no, because "elemental lithium is the same element whether it is in the earth or isolated." That would also seem to be true of a gene whether it is in the body or isolated, as the dissent pointed out; the decision doesn't elaborate on where it sees a difference.

The partially concurring opinion also stumbles on this point. In it the judge concludes that short fragments of the gene with a clear purpose are probably patentable on the grounds of their utility. Larger fragments of a gene, however, are too big to be useful, and more closely resemble the gene's natural state. So, beyond some completely undefined size point, the judge would probably rule that genes cannot be patented. The only thing that prevents him from doing so is the long history of allowing gene patents.

The dissent is critical of the patents in other ways. For example, it notes that the full sequence of BRCA wasn't even known when the patents were filed; large stretches of sequence were simply filled in with a generic symbol instead of the usual As, Tc, Cs, and Gs. "An almost incalculably large number of new molecules could be created by filling in those gaps with almost any nucleotide sequence," judge William Bryson wrote, "and all of those molecules would fall within the scope [of the patent]."

The dissent also notes that the allowing the patent to cover the isolation of the BRCA gene raises questions about whether anyone that sequences a human genome would be violating it. Necessarily, that process would involve the purification of the relevant DNA; the only thing it might not involve is the specific analysis of the BRCA sequences, but the ruling had already thrown those patents out.

So although the Myriad patents will stand for now, it's clear that there's not unanimous agreement about what, precisely, is appropriate for patenting. Which raises the prospect that a different legal approach, focusing on different patents, might get a more favorable reception.