Experimental Use in Plant Sciences Research

The patent infringement provision in US law speaks in absolute terms. The provision prohibits unauthorized ‘making’ or ‘using’ of a patented invention, without regard for whether the alleged infringement is innocent, de minimis , or undertaken in an arguably non-commercial setting (35 U.S.C. §271(a) ). The absolutist approach of the patent infringement provision offers some social benefits: it may reinforce the innovation incentives that the patent grant is thought to provide, and may facilitate market transactions in patent rights by eliminating a source of uncertainty over enforcement. On the other hand, the absolutist approach also imposes social costs: it may chill follow-on innovators, even those who are exploiting the patented technology for purposes of academic research.

When confronted with these competing considerations, the US Congress and courts have only rarely departed from the absolutist structure of the US patent infringement provision. Where alleged infringers have claimed that their unauthorized ‘use’ of a patented invention was in the course of experimentation, US courts have only grudgingly recognized, and almost never applied, an exception to infringement liability, characterizing the experimental use exception as ‘truly narrow’ ( Madey v Duke University , Roche Products v Bolar Pharmaceutical ). Congress has crafted a statutory exemption that is limited to experimentation ‘solely for purposes reasonably related to’ the development of data for submission to the Food and Drug Administration (FDA) (35 U.S.C. §271(e)(1) ). The US Supreme Court has held that the §271(e)(1) exemption appplies to ‘the use of patented compounds in preclinical studies . . . as long as there is a reasonable basis for believing that the experiments will produce’ the type of data that would be relevant for an eventual FDA submission ( Merck KgaA v Integra Lifesciences I ).

Patent scholars have long been intrigued by the experimental use doctrine, and have exhaustively explored its policy implications (e.g. Hantman, 1985; Eisenberg, 1989; Karp, 1991; O’Rourke, 2000; Cohen and Lemley, 2001; Mueller, 2001; Strandburg, 2004). However, despite numerous calls for the creation of a regime of formalized experimental use rules (including some recent debate about whether to expand the existing §271(e)(1) exemption to cover plants), the US Congress has declined, to date, to implement experimental use legislation. By contrast, in Europe, a generic experimental use exception has been codified in many national patent statutes, and in Germany, a particularized experimental use exception directed at shielding experimental plant breeding from patent liability has been incorporated into the patent statute (Deustcher Bundestag, Drucksache, 2003).

The basic policy arguments underlying the experimental use doctrine are well understood. A robust experimental use exception is thought to lower the cost of researchers conducting follow-on research, but is also thought to reduce patent royalties (by shielding research uses from liability) and shorten the effective patent term (by reducing the costs to researchers of designing around). The net effect on incentives to invest in innovation is not clear, and even if it were, it is not clear that actual researcher behaviour would approach the rational instrumentalism that this simple cost–benefits analysis might imply. Accordingly, the way forward for the application of the experimental use doctrine in modern US patent law, and particularly in patent law concerning plant sciences, is not clear.

The debate over the experimental use exception resembles the debate over application of other claim scope doctrines such as enablement. Like enablement, experimental use appears to be operating more like a standard than a rule. Experimental use appears to be largely characterized by ex post evaluations of conduct, except that although ex post evaluations in enablement are undertaken by judges in contested litigation, ex post evaluations of experimental use in plant sciences (and in many other areas) are apparently being undertaken informally, by parties operating in the shadow of potential litigation.

Perhaps predictably, many suggestions for reforming experimental use are directed at attempting to transform experimental use into a formalized hierarchy  of rules. Many of these reform arguments focus on the need to create a sharply defined safe harbour for follow-on researchers, reducing the costs imposed on them by the threat of patent litigation. However, for many of the same reasons that I offered in connection with enablement, the costs of attempting to formulate a code of rules for experimental use might swamp any savings realized.

This brings us again to the question of whether accepting experimental use as a standard, rather than as a rule, is tantamount to remaining passive while the doctrine merely runs its course in the various technology areas in which it is important, including the plant sciences. The answer, again, is ‘no’. Here, as with enablement, efforts could be directed towards reducing the cost of the information necessary to make correct decisions, rather than being directed towards formulating new rules.