Ben Goldacre, The Guardian, Saturday 2 April 2010
This week the Association for Molecular Pathology, working with the American Civil Liberties Union, won a major victory, overturning just some of the patents owned by a company called Myriad on the BRCA1 gene for breast cancer. There are three reasons why gene patents like these are stupid: only the last one is funny.
Patents in general are a sensible idea, because people are more likely to invest in innovation if they believe it will give them a competitive advantage over other people, and because patents allow people to share their discoveries safely, instead of monetising their advantage by keeping a discovery secret. But patents also act as a barrier to innovation, and gene patents bring these disadvantages into stark relief.
Different people have slightly different forms of the BRCA1 gene, and these confer different risks for breast cancer, so doctors like to run tests and see which form you have. Myriad were not granted a patent on these tests: instead they got a patent on the BRCA1 genes themselves, which are out there, present in humanity, and naturally-occurring parts of our bodies. This has had a chilling effect on clinical activity and research.
One of the plaintiffs in the case against them, for example, is a patient who had a BRCA1 test from Myriad, and would like it independently verified by someone else’s test of the same gene. She cannot have such a test: in the US, only Myriad are allowed to offer BRCA1 tests (and they charge over $3,000). The company have gone after people developing tests for risk of cancer using the BRCA1 gene, and this has retarded the development of new tests.
In fact, a survey in 2003 of all leading laboratory directors in the US looked at the extent of this research chill in all areas of medicine from gene patents, and found that 53% had “decided not to develop or perform a test/service for clinical or research purposes because of a patent”. This is not surprising, since a study from 2005 found that about a fifth of all the 23,688 genes in the human genome – the code that makes up you – have already been patented.
But these tests are just the tip of the research iceberg. Almost all basic science research on the BRCA1 breast cancer gene over the past 12 years has infringed Myriad’s patent, and although the company has tended not to go after basic science researchers, they have never promised that they won’t in the future, so this academic research on a major risk factor for a major killer – the most common cancer in women worldwide – continues only with Myriad’s indulgence, making it risky and uncertain work.
But that’s only the half of it.
A paper titled “Metastasizing patent claims in BRCA1”, just published in the journal Genomics, examines the true extent of the BRCA1 patents granted in 1998, and they are laughably absurd. The authors examine patent #5,747,282. This makes claim to any sequence of 15 nucleotides, the “letters” of the genetic code, coding for any part of the protein made by the BRCA1 gene (and if you can’t understand that last sentence, then you’re as ignorant as someone who has never read Coleridge).
First they calculate how absurdly broad this claim is, from first principles. There are about 1.6 million different 15-nucleotide sequences that could code for some part of the BRCA1 protein. There are 1.07 billion possible different 15-nucleotide sequences in total. Therefore, this patent covers roughly one in every 600 of all imaginable 15-nucleotide DNA sequences. Since a typical human gene is a sequence of about 10,000 nucleotides, then on average (if you assumed that human genes were random strings of nucleotides) you would expect every human gene to contain about 15 of the 15-nucleotide sequences claimed under the BRCA1 patent.
Then they tested their model against reality: in a giant computing task, they took all the 15-nucleotide sequences from the BRCA1 gene, and searched for them, just on chromosome 1: they found 340,000 matches, roughly the same as their theoretical prediction, and the equivalent of 14 infringing sequences on every human gene. The BRCA1 gene, incidentally, is on chromosome 17.
The claims in this patent therefore extend, if properly enforced, to almost every single gene, in every single person on the planet. There is a moral and practical argument to be had about patenting nature, but the rights conferred in this patent are basically absurd.