By Kevin E. Noonan —

There was a time not so long ago where the U.S. Patent and Trademark Office seemed to be sincere in trying to accommodate the needs of inventors and the public in providing fair examination of cutting-edge technologies.  This is particularly true in the biotechnology arts, with such innovations as the sequence listing database and (after some resistance) recognizing that nucleic acid claims were not prima facie obvious.  That has changed, with the Office seeming to be happy in many instances to play an advanced game of "gotcha" with its biotechnology customers.

This new attitude was illustrated once again in a presentation by Julie Burke, a "Quality Assurance Specialist" in Group 1600 at the Biotechnology/Chemistry/Pharmaceutical Customer Partnership meeting held at the Patent Office this Wednesday.  In her presentation she addressed "incorporation by reference," the sufficiency of disclosure requirements, and the standard for determining whether clarification of the identity of certain nucleic acids would be new matter.  The particular issue involved whether and under what circumstances it was sufficient to identify a nucleic acid by its GenBank Accession Number.  Ms. Burke sets out the "problem" in the very first substantive slide:

In patent specifications, every element or ingredient of the product should be set forth in positive, exact, intelligible language, so that there will be no uncertainty as to what is meant.

The relationship between sequence information submitted to a public database and the amino acid or nucleic acid product it identifies is sometimes indefinite, uncertain, and arbitrary.

She goes on to identify the bases for the "uncertainty":  sequence information in public databases can "change from time to time," and these changes can be made by third parties.  These changes can constitute new matter, according to Ms. Burke, and implicate the Office’s "New Matter Guidance" to Examiners to sensitize them to the issue.  These considerations implicate the requirements for incorporating material by reference, which also is subject to new matter prohibitions, i.e., sequences that have been changed after an application incorporating them by reference has been filed.  The rationale is that only a sequence having one version has been "uniquely identified"; even prior to the filing date, when multiple sequences for the same gene are identified, the existence of these sequences means the gene "may not be uniquely identified."

Many of the examples in Ms. Burke’s presentation involve claims to isolated nucleic acids, and insofar as the application of the "guidelines" are limited to such claims they may have some justification.  But the reality is that the 25,000-30,000 human genes found by the Human Genome Project have been disclosed in public databases for almost 10 years, and patent applications on these genes were filed at the turn of the century.  Ms. Burke’s presentation also mentions that the Office has changed its former policy of examining up to 10 sequences in a single application, and will now only examine one sequence per application (OG Notice dated March 27, 2007; see Patent Docs report).  As a consequence, only a small percentage of human genes have been patented to date, and the likelihood is that the 20-year patent term limits will preclude most of the genes disclosed in most of the applications from ever being the subject of granted claims.  Moreover, the business and intellectual property of the next 50 years in biology will not be the identification of genes but rather how different genes are differentially expressed in disease (inter alia, for diagnostics) or in response to different physiological stimuli such as therapeutic drug administration.  These experiments take advantage of the information about these genes, using array technologies and other methods.  The issue thus is not the same as with claims to the genes themselves; all that is required is that the skilled worker understand which genes are differentially expressed.  Sequence variations, whether the results of allelic variation or sequencing disparities from different research groups thus would not create any ambiguity for the skilled worker as to the identity of the genes.  However, the current Patent Office focus on whether there are sequence variations that preclude these genes from being "uniquely identified" elevates the formalism of sequence identity (important for sequence claims) to another "bright line" rule having the potential for being indiscriminately applied.  And the proposed solution, including the sequence of genes in a sequence listing, merely increases the applicant’s burden without improving the quality of the disclosure for known genes.

Which, unfortunately, is not a unique position for the Office to take in recent years.  Perhaps a change of administration will return the Office’s attitude from confrontation back to cooperation.