Ernest I. Gifford and Avery N. Goldstein
8.3 Information needed to draft a patent application
The inventor needs to provide enough information so that the attorney or agent can describe the invention in enough detail so that ‘‘one of ordinary skill in the art’’ can make and use the invention. Patent law also requires that the patent application reveal the ‘‘best mode’’ of making and using the invention, if work on the invention has progressed far enough so that a
‘‘best mode’’ is known.
The inventor needs to provide the attorney with information on how the invention is distinguished from, or better than, devices or procedures in the prior art. However, the attorney should refrain from including this distinguishing information in the patent application itself, for reasons de- scribed below. After the patent application is filed, and where several months go by, and the examiner mails an Office Action to the attorney, the attorney will need the distinguishing information to rebut 102-rejections and 103-rejections contained in the Office Action.
The attorney should refrain from spontaneously writing this distin- guishing information in the patent application, or in other documents, such as an Information Disclosure Statement or a Petition to Make Special, as described in Gentry Gallery v. Berkline Corp. (Fed. Cir. 1998) [50]. State- ments to the effect that one’s invention has specific properties that overcome problems found in the prior art can reduce the scope of coverage of the claims, as illustrated by Dawn Equipment v. Kentucky Farms(Fed. Cir. 1998) [51], a case concerning farm equipment. Dawn Equipment’s device used a
GOLDSTEIN/Patent Law for Scientists and Engineers DK2550_c008 Final Proof page 166 3.2.2005 2:36pm
166 Tom Brody
rotating pin, while Kentucky Farms’ device used an inserted pin and multiple holes. Dawn Equipment, the patent owner, had a good chance of winning the lawsuit, but for one type of comment in its patent. That comment was a series of criticisms of multihole devices of the type made by the competitor, Kentucky Farms. This written criticism of multiholed devices occurring in Dawn Equipment’s patent prevented this patent from covering the multiholed device made by Kentucky Farms. A similar lesson was provided by Signtech USA v. Vutek(Fed. Cir. 1999) [52], a case concerning ink-jet printers. Another example can be found inGentry Gallery v. Berkline (Fed. Cir. 1998) [50]. The statements written by the attorney during the prosecution of Gentry Gallery’s U.S. Patent No. 5,064,244 were that ‘‘the invention as described is the only possible configuration of the invention’’ and that ‘‘variations of the described invention are outside the purpose of the invention.’’ Unfortunately for Gentry Gallery, Inc., these statements prevented it from asserting its patent to prevent a competitor from making sofas, where these sofas were variations of Gentry Gallery’s sofa.
8.3.1 Examples where not enough information was provided
The inventor needs to tell the patent attorney or agent how to make and use the invention. Where the inventor knows how to make the invention, but willfully refrains from revealing a step (especially a nonroutine step) to the attorney, the patent can be rejected for lack of enablement. Rejections for lack of enablement are under (based on) 35 U.S.C. 112. Chiron v. Abbott (D.N.Cal 1995) illustrates the situation of U.S. Patent Application No.
659,339, filed October 10, 1984, where the application failed to provide a description that enables the invention. The application concerned an assay method for AIDS virus [10, 11]. The assay method requires a purified viral protein (envelope protein). Envelope protein is not commercially available.
Envelope protein must be produced using a clone of some DNA coding for the envelope protein. The inventor’s patent application failed to reveal the source of the cloned envelope protein.
The court pointed out that the inventors had published a paper in the journalNatureprior to the filing date, revealing how to clone the envelope gene. The court pointed out that the inventors failed to cite this publication in their patent application. The court pointed out that the inventors had cloned the envelope protein (a good thing), had deposited the cloned envelope protein in a public bank (a good thing), but that the patent application failed to mention this deposit (a bad thing). The publication in Nature, as well as the date of deposit, took place before the October 10, 1984 filing date, and hence could easily have been mentioned in the patent application. The claim was invalidated under 35 U.S.C. 112 [10, 11].
The following example concerns an invention of a recombinant plant.
Enzo is the owner of U.S. Patent No. 5,190,931. The example is reported
GOLDSTEIN/Patent Law for Scientists and Engineers DK2550_c008 Final Proof page 167 3.2.2005 2:36pm
Chapter eight: Interactions with a patent agent or attorney 167
inEnzo v. Calgene(D. Del. 1998) andEnzo v. Calgene(Fed. Cir. 1999) [32, 53].
Dr. Inouye, the inventor in the patent, succeeded in using antisense tech- nology for regulating three genes in prokaryotic cells (in Escherichia coli).
The inventor’s patent application described three working examples from E. coli, but failed to mention the inventor’s nonworking examples in attempts in eukaryotes (yeast).
The inventor succeeded in persuading the examiner that the antisense method could be applied in eukaryotic cells, and that the method could be accomplished without an unusual degree of experimentation. To per- suade the examiner, the inventor filed a Declaration from an independent expert, stating that Dr. Inouye’s invention could be applied without undue experimentation to all genes in all cells.
The patent was then issued as U.S. Patent No. 5,190,931. The patent has a claim that reads, ‘‘A prokaryotic or eukaryotic cell containing a . . . DNA construct, which . . . produces an RNA (antisense RNA) which regulates the function of a gene. . . . ’’
During a subsequent lawsuit, a number of problems began to unfold for Enzo and its patent:
1. It was revealed that the inventor had stated in a grant applica- tion that ‘‘there are no consistent ways in which to reliably generate an effective antisense gene for use in the inhibition of a . . . gene.’’ The grant application was dated April 1, 1990, well after the date of filing the patent application (October 20, 1983).
Furthermore, the inventor also stated in an article, published on June 15, 1988, that there was no success with yeast (a eukaryotic cell). Again, this date was after the patent’s filing date (October 20, 1983), indicating that these problems also existed at all earlier times;
2. Enzo’s patent claimed a method in prokaryotic cells, and contained three examples from prokaryotic cells, but it claimed a method in eukaryotic cells, with no examples; and
3. A witness at the trial stated that trying to make the invention work required ‘‘witchcraft’’ and thus could not have been enabled by the specification.
The above three problems all point to one conclusion — that Enzo’s patent application failed to enable the invention in eukaryotic cells. Rejec- tions due to lack of enablement will occur where the claims describe an invention where a working model has already been made in the laboratory, but was not adequately described in the patent application. Rejections due to lack of enablement will also occur where the description was overly predictive and where the invention could not possibly have been made in the laboratory at the time of filing.
GOLDSTEIN/Patent Law for Scientists and Engineers DK2550_c008 Final Proof page 168 3.2.2005 2:36pm
168 Tom Brody
The following test lends perspective to the above fact pattern. The court applied the Wands factor test [54]. This test involves comparing the follow- ing factors:
1. The amount of experimentation required, i.e., whether it is more than a routine amount of experimentation. This factor does not relate to whether an experiment requires an entire year or longer
— it relates to whether the time is typical for a worker doing that sort of experiment;
2. The amount of direction or guidance needed;
3. The presence of working examples;
4. The state of the prior art (whether there are many published articles on how to make related inventions work);
5. The predictability of the art — the concept of predictable and unpredictable fields has been discussed [55]; and
6. Whether the claim in the patent claims a great deal more than what is described by the examples in the specification in the patent application, or if the claim only claims what has been shown in the examples.
The court pointed out that Enzo’s patent claimed more than what had been described in the working examples, that the claim related to an unpre- dictable art (biology), and that the working examples were only fromE. coli (and not from eukaryotes). In short, Enzo’s claim failed the Wands test. The court invalidated Enzo’s claim under 35 U.S.C. 112.
The take-home lesson is that any step in making an invention that cannot be accomplished by known, routine methods should be described in the patent application to the point where one of ordinary skill in the art can make and use the invention.
8.3.2 The inventor needs to provide information on the best way (best mode) of making and using the invention
The patent application must reveal the best way to make and use the invention. Where this requirement is not met, the patent application will be in violation of 35 U.S.C. 112. The best mode is illustrated byEli Lilly Corp.
v. Barr Laboratories (Fed. Cir. 2000), in the Case Studies [56]. If a patent application does not reveal the best mode, and the examiner fails to make a 112-rejection and allows the patent, the patent is likely to be invalidated during a subsequent lawsuit. This invalidation will be under 35 U.S.C. 112.
A question related to satisfying the best mode requirement is, what if one of the examples in the specification section does not work at all, i.e., what if the ‘‘worst mode’’ is inadvertently described? The answer is that a patent will probably not be invalidated if one or two of the examples do not work, as long as a greater number of working examples are described [57].
GOLDSTEIN/Patent Law for Scientists and Engineers DK2550_c008 Final Proof page 169 3.2.2005 2:36pm
Chapter eight: Interactions with a patent agent or attorney 169
8.3.3 The inventor should provide the attorney with information regarding the level of skill of one of ordinary skill in the field
The inventor needs to provide information on the level of skill of ‘‘one of ordinary skill in the art,’’ so that the attorney can write claims that only cover versions of the invention that an ordinary worker can figure out from reading the specification. In other words, the detail in the specification needs to be aligned with the level of skill of an ordinary practitioner in the relevant technology. More detail may be provided, but less detail will invite a 112-rejection. In evaluating patent applications (or patents), the examiner (or court) will use this level of skill as a set point, or balance point, for comparing the ‘‘mass’’ of the specification with the ‘‘mass’’ of the claims.
Amgen v. Hoechst and Transkaryotic Therapies(D. Mass 2001) provides an illustration of the concept of one of ordinary skill in the art [58]. In U.S.
Patent No. 5,955,422, Amgen disclosed examples of making the invention (erythropoietin), where one example was with monkey cells and another example was with hamster cells. Examples with other mammalian cells were not shown. Amgen’s Claim 1 contained the term ‘‘mammalian cells.’’
Transkaryotic Therapies wanted to persuade the court to invalidate the claim, and thus argued that the specification section in Amgen’s patent did not contain enough information to guide the ordinary laboratory worker to make the invention in other types of mammalian cells. Transkaryotic Therapies’ goal in making this argument was to persuade the court that Amgen’s patent should be invalidated for failure to satisfy the enablement requirement (35 U.S.C. 112). Amgen successfully defended its claim by providing an expert, Professor Harvey Lodish, who stated that ‘‘it should be easy to figure out experimentally’’ how to produce erythropoietin in a variety of other mammalian cells.
Another example is mentioned but not discussed. In a lawsuit involving construction materials, the court considered whether one of ordinary skill in the art was ‘‘Barney the BrickLayer’’ or one ‘‘familiar with the literature in the field [who] might be expected to supervise building-reinforce- ment crews and would be likely to attend the World of Concrete Trade Show’’ [59].