Reply to "Impact of gene patenting on R&D and commerce"

Sandy Thomas responds

Stott and Valentine are right to point out that all patent statistics should be treated with caution. It is widely recognized that patent data, as partial scientific indicators, are only proxy measures that may provide some insights into the process of innovation¹. As indicators, they feature immense variability depending on, for example, the subject matter and the scope of the claims, as well as the context, including the speed of examination and the national or regional regulations. Patent databases as a consequence are inherently variable; in addition, companies are acquired and de-merged, complicating the picture. The role of patents in innovation is complex, poorly understood and not at all well supported by empirical evidence². But a proxy measure is better than no measure at all.

The GENESEQ data tell us only about applications not granted patents, but they also only give information on one member of each patent family (the 'basic' patent application). This kind of (essentially bibliometric) research can only ever be a proxy measure of downstream intellectual property rights activity. Our approach represents an attempt to add a large-scale, international analysis to a meager literature base. It is also complementary to the case study approach that others have undertaken.

Although we agree that data based on granted patents can be informative about what is actually protected, we would argue that data on filed patents can nevertheless inform us about something rather different: namely, how patent applicants set out to protect their inventions at a relatively early stage of the innovation process. The results show a very substantial intent to claim patent rights. They also identify major players at the application stage. Not all patent applications will be granted, but it is fair to say that in the United States there has tended to be a presumption to grant patents and that the US Court of Appeals for the Federal Circuit has been highly protective of patents in the biotechnology sector. Even data on granted applications have their limitations: only a small percentage will actually be used in the marketplace. This is what we should expect: that the higher levels of early protection on promising inventions will diminish such that only the fraction which have commercial potential are given sustained protection.

We also agree that the patent regimes in the United States and Europe have differed with respect to the patentability of DNA sequences. Nor are they static. As we pointed out in our article, greater stringency can be expected in the light of the US Patent and Trademark Office Utility Guidelines³ adopted in 2001, which have in effect been endorsed by the European Patent Office⁴. To what degree patents filed between 1996 and 1999 will be granted is uncertain. But past trends in the United States suggest that many applications may be successful. One may reasonably speculate that a trend towards more stringent examination may, inter alia, discourage filing in the future.

The question of the lack of a statutory research exemption in the United States is an important one. Stott and Valentine suggest that the lack of an exemption, and its presumed negative effects, has not prevented the United States from being the most innovative country in the world. But how much of this innovation depends on what some hold privately to be large-scale infringement of patents by researchers across the United States? If the Madey v. Duke decision (Nat. Biotechnol. 21, 453, 2003) is confirmed by the US Supreme Court, it is a distinct possibility that some infringements, particularly in universities, may be pursued. The case for a statutory research exemption or a similar instrument to protect the freedom of researchers to undertake pre-competitive research without licensing in patented inventions might then become compelling.

See "Impact of gene patenting on R&D and commerce" by Stott et al.