Enablement of Transgenic Plant Inventions

A US patent document must provide a description of the patented invention that is adequate to enable a person of ordinary skill in the art to make and use the claimed invention (35 U.S.C. §112, 1st). The enablement requirement seeks to ensure that patentees provide high-quality teachings that correlate in scope with the scope of the potentially valuable patent rights that they receive; as such, enablement is an essential part of the quid pro quo of the patent system (in re Wright, Enzo Biochem, Inc. v Calgene, p. 1561). Courts have established that a patent document satisfies the enablement requirement even if ordinary artisans would need to undertake some experimentation in order to make or use a patented invention, as long as experimentation is not deemed to be 'undue' (in re Vaeck). Courts have to balance a number of factors when determining whether experimentation is undue for enablement purposes, including the often critical factor of predictability in the art (in re Wands). The Federal Circuit has identified biotechnology as one of the unpredictable arts for enablement purposes, though it has also recognized that assessments of predictability should be revised as science advances (in re Wright, Enzo Biochem, Inc.). Relatedly, the Federal Circuit observed, in a biotechnology enable-ment case, that the enablement requirement should be applied with particular vigour in 'nascent technologies' (Chiron v Genentech). Courts have also held that the enablement inquiry must be undertaken in view of the state of the art as it existed at the time when the inventor filed the patent application (in re Goodman).

An issued patent is presumed to comply with the enablement requirement, a consequence of the general presumption of validity accorded to all issued patents (35 U.S.C. §282). Patent challengers bear the burden of proving a failure of enable-ment, and must establish the relevant facts by the standard of clear and convincing evidence in order to overcome the presumption of validity (Chiron, p. 1252).

In a series of cases, the Court of Appeal for the Federal Circuit (the most important US court on patent law matters) has applied the enablement requirement to transgenic plant inventions. In Adang v Fischoff, an appeal from an interference proceeding, in which two inventors, Adang and Fischoff, both claimed rights to the same invention, the court analysed the validity of the following interference count:

A tomato plant which has been regenerated from a tomato plant cell transformed to comprise a full length Bacillus thuringiensis crystal protein gene capable of encoding a Bacillus thuringiensis crystal protein of about 130 kD under control of a promoter such that said gene is expressible in said plant in amounts insecticidal to Lepidopteran insects.

Adang's patent specification included an example reporting the transformation of tobacco cells with a Bt gene having the characteristics called for in the count, and a list of 94 types of plants (including tobacco and tomato) described as being capable of being transformed by Bt crystal protein genes.

The court concluded that the disclosure did not enable the count. References from the technical literature, even after the filing date of Adang's application, showed that attempts to incorporate a full-length Bt crystal protein gene into various strains of tobacco plants using Agrobacterium-mediated transformation failed to produce stable transformations or predictable insect toxicity - and in some instances failed even to produce viable plants. Although both Adang and Fischoff had submitted expert evidence (testimony and new experimental evidence), Fischoff's was more credible; Adang's seemed to be based on hindsight. The court upheld the conclusion of the US Patent and Trademark Office (PTO) that it would have required undue experimentation on the part of a worker of ordinary skill to use the Adang disclosure to transform and produce a transgenic tomato plant within the scope of the interference count (Adang, pp. 1355-1358).

In Plant Genetic Systems v DeKalb Genetics, the patent in suit claimed a transformed 'plant cell':

1. A plant cell having a heterologous DNA stably integrated into its genome; said DNA comprising a heterologous DNA fragment encoding a protein having an acetyl transferase activity which inactivates a glutamine synthetase inhibitor in said cell.

The patent disclosed working examples involving the Agrobacterium-mediated transformation of tobacco, tomato, sugarbeet and potato plants - all dicots. Yet the term 'plant cell' in claim 1 arguably covered not only dicots, but also mono-cots (such as DeKalb's transgenic corn varieties). Likewise, claim 1 arguably covered any transformation technique. The enablement issue was whether the disclosure would have enabled a person of ordinary skill in the art to employ any transformation technique, including Agrobacterium-mediated transformation, to transform any plant cells, including monocot plant cells, as of the 1987 application filing date.

The trial court concluded that the patent specification did not provide an enabling disclosure of claim 1, and the Federal Circuit upheld this judgment on appeal. The trial court relied on a prior case, in re Goodman, and a detailed analysis of the technical literature and expert testimony. Concerning Goodman, the trial court had pointed out that the facts of record there supported the proposition that a viable Agrobacterium-mediated transformation technique for monocots was not generally known in the art as of 1985. The trial court concluded that Goodman therefore supplied a starting point for the trial court's enablement analysis; the trial court could look at the literature between 1985 and 1987 to determine whether the transformation art had advanced during that time period. The Federal Circuit approved of this methodology, rejecting the argument that the trial court had effectively shifted the burden of proof on enablement to the patentee.

Although the technical literature had reported one transformation of corn by Agrobacterium in 1986, other contemporaneous literature references expressed reservations about this report, as did the patentee's own personnel. Early 1990s work reported only limited success in transforming corn by electroporation and microprojectile bombardment, and not until 1995 did the patentee succeed in transforming corn via Agrobacterium. The Federal Circuit concluded that the trial court had properly considered the reports of successes in the 1990s in the enable-ment analysis because they raised the inference that undue experimentation would have been required to achieve success as of 1987, the relevant date for the enablement inquiry.

In a related litigation, the enablement issue came before the Federal Circuit again. In Monsanto v Bayer Bioscience, Monsanto (successor to DeKalb Genetics) sued Bayer (successor to Plant Genetic Systems) for a declaration that several patents relating to Bt technology were not infringed and/or were unenforceable. The patents claimed methods for transforming plants with a truncated Bt gene (and related genes and transformed plants and plant cells), with no express limitation to particular plants. Yet the patents apparently disclosed only the transformation of tobacco plants, again presenting the question of whether a claim encompassing all plants, including monocots, was enabled by a specification that disclosed only the transformation of a dicot, in view of the state of the art in 1986, the relevant application filing date.

The trial court undoubtedly would have been correct to look to prior cases for factual support for the proposition that the disclosure in the patents did not enable the claims, but it used a different strategy, concluding that the patentee Bayer was precluded even from litigating the issue, on the ground that Bayer's predecessor had litigated and lost a similar issue in Plant Genetic Systems. The Federal Circuit reversed. The patent disclosure at issue was not identical to the disclosure that had been in issue in Plant Genetic Systems. While Bayer was bound by the prior finding that the Agrobacterium-mediated transformation of monocots was not established in the art as of 1986, Bayer was still entitled to litigate the question of whether its patent disclosure supplied the missing information.

These cases are related in some detail to illustrate a broader point about the character of second-stage patent jurisprudence. None of these enablement decisions articulate especially bold new statements of the law. Instead, they offer insights tied closely to the particular facts of the cases. Monsanto in particular exemplifies the rather extreme fact specificity of the enablement inquiry. Each of the cases turns on close assessments of the technical literature and credibility judgments about conflicting expert testimony.

On first glance, these cases may seem to refute the claim that the enablement doctrine could serve as an important policy instrument in shaping patent rights in plant sciences. If policy advancement resides in the promulgation of a constellation of new, bright-line rules, the pessimistic assessment of enablement's potential may be correct. These cases surely illustrate that enablement is not a likely platform for the development of rule-bound law.

However, we should resist the notion that policy advancement in patent law requires the articulation of new catalogues of special rules. We should also resist the quick dismissal of the enablement doctrine as a policy tool. The very essence of 'second-stage jurisprudence' is that policy advancement can occur: (i) incrementally; and (ii) by way of a few flexible standards rather than a multitude of intricate rules.

These observations connect to a larger jurisprudential debate between the relative merits of rules versus standards. A full consideration of this debate, even limited to its application in patent law, is beyond the scope of this chapter. However, a brief introduction is pertinent here. A 'rule', as understood in this debate, involves a high degree of ex ante determination of authorized (or required) conduct, leaving limited discretion to the adjudicator, while a 'standard' involves a high degree of ex post determination, leaving to the adjudicator the bulk of the task of determining authorized or required conduct and adjudicating disputed facts (Kaplow, 1992). Actual legal commands cannot be expected to organize themselves neatly into one or the other category, but it is plausible to speak of a continuum of legal commands running from rules at one extreme (constraining judicial discretion) to standards at the other (furnishing discretion) (Lee, 2002).

Whether a given legal problem calls for resolution by a rule or by a standard is a complex judgement, but one simple way to think about the problem is to consider a rough balance of costs and benefits. A rule is relatively costly to promulgate but should be relatively cheap for judges to apply (and for interested parties to apply when attempting to predict outcomes). A standard is relatively cheap to promulgate but may be costly for judges to apply (and costly for interested parties, because outcomes may be less predictable) (Kaplow, 1992, p. 621).

In the lexicon of rules versus standards, enablement currently operates as a standard, as Adang, Plant Genetic Sciences and Monsanto illustrate. The enablement requirement does not particularize in advance the precise level of disclosure that will be compliant. Instead, it states a generalized aspiration and leaves PTO examiners and judges broad discretion to adjudicate enablement ex post. This approach entails certain costs. As the plant sciences enablement cases illustrate, enablement is costly for litigants (and judges) to apply, as it requires intensive development and evaluation of technical evidence. Outcomes in enablement cases, including those in the plant sciences, have been criticized as unpredictable (Todaro, 1994). Moreover, the fact that the enablement determination is made ex post, sometimes years after the application filing date, yet must evaluate the disclosure in view of the state of the art existing as of the application filing date, creates the potential for error through reliance on hindsight.

These costs are significant, and these types of arguments about costs are standard fare in patent policy circles. In particular, arguments about the value of predictability in patent rights form the centrepiece of much of the Federal Circuit's recent thinking about patent scope, on issues such as claim interpretation and limits on the doctrine of equivalents.

What is routinely missing from these debates over claim scope is an analysis of the other side of the balance, i.e. a recognition that the process of formulating detailed rules on enablement and other claim scope doctrines would likewise entail significant costs. At this early stage in the history of utility patents for plants, there is a lack of sufficient information to design particularized enablement rules for plant sciences innovations. The enablement cases discussed earlier i ndicate that the information set necessary for designing any such rules is a complex one.

In addition, given the rapid pace of technological change, and change in our understanding of the innovation process, any particularized rules would quickly become obsolete. These sorts of objections - information costs, obsolescence - are commonly offered against rule-bound law (Sunstein, 1995).

In view of these countervailing considerations, policymakers should resist the temptation to craft (or attempt to craft) particularized, bright-line rules to govern matters of patent scope in the plant sciences. At least in this second stage of jurisprudence, policymakers should allow doctrines like enablement to operate as flexible standards, on the grounds that it is likely to be more efficient to develop the law incrementally, through case-specific judgements, than to engage in a complicated rule-making exercise.

This line of argument has its limits; it does not advocate untrammeled judicial discretion over claim scope determinations. Consider, for example, the US patent law's 'written description' requirement. The Federal Circuit has held that 35 U.S.C. §112, para 1 imposes a requirement that the patent applicant provides an adequate written description of the invention in addition to providing an 'enabling' description of the invention (University of Rochester v G.D. Searle). Many have criticized the court's written description jurisprudence as essentially standardless, i.e. as not even articulating a discernible standard that supplies minimal guidance for the exercise of judicial discretion (Janis, 2000). Here, the costs of unpredictability are substantial, and the cost of articulating at least a flexible standard is surely not so high as to justify the current, standardless approach.

To a reformer, the suggestion that the enablement doctrine be allowed to operate as a standard, and that it be left to evolve incrementally through case law may seem too passive. But this is deceptive. It is quite possible to fashion policy initiatives proactively in a standards-driven environment. For example, there are several ways in which the enablement standard, as applied to plant sciences inventions, might be enriched without converting it into a bright-line rule:

1. Develop a better understanding of the qualities of the 'person of ordinary skill' in the plant biotechnology art.

2. Develop a better fact base for assessing whether particular endeavours within the plant sciences are 'predictable'.

3. Continue to develop a fact base for assessing undue experimentation.

One mechanism that might be considered for implementing these suggestions is the process of developing examination guidelines at the PTO. The process of formulating examination guidelines for enablement in plant biotechnology might yield benefits even apart from the substantive content of any such guidelines. The guidelines drafting exercise could serve as a forum in which researchers can inform the PTO about relevant facts within the plant sciences industry. In addition, because guidelines are less formal than legislation or even regulations, guidelines may in theory be formulated more rapidly, and may thus prove to be a good mechanism by which the PTO can respond to rapid shifts in the technological landscape.

Similarly, because guidelines need not be permanent, they can provide a good forum for policy experimentation. Mistakes can be more readily addressed, and hypothetical illustrations can be explored and analysed. There is some evidence that examination guidelines can operate constructively in the patent system, and can inform judicial decisions as well as PTO practice (e.g. in re Brana, Enzo Biochem, Inc. v Gen-Probe, Inc.; Utility Examination Guidelines, 2001).

Examination guidelines also present significant potential downsides. Because they do not have the force of law, they are inherently less capable of reducing uncertainty costs than would a binding set of legal commands. In addition, a guideline drafting exercise always carries the risk of devolving into an exercise in formalism (e.g. Pila, 2003). An effort to develop bright-line rules in the guise of guidelines would be as counterproductive as an effort to develop bright-line legislation for enablement.

0 0

Post a comment