* The content of this article was presented at the 16th Annual ATCC Biotech Patent Forum on September 25, 1998.

** The authors would like to thank Jorge A. Goldstein, John M. Covert, and Robert W. Esmond for their participation in valuable discussions concerning the topic of this paper.

*** This article reflects the present thoughts of the authors, and should not be attributed to Sterne, Kessler, Goldstein & Fox P.L.L.C. or any of its former, current, or future clients.

^©1998, Sterne, Kessler, Goldstein & Fox P.L.L.C.

I.       Introduction

            The benefit of having an earlier filing date than the competition usually warrants filing a patent application soon after an initial discovery is made. However, due to the basic nature of Biotech research, an initial discovery is often less important commercially than later discovered improvements. When developing strategies to maximize patent protection, the interplay of patent issues for inventions resulting from the initial discovery and for inventions resulting from the improvement must be considered.

            For example, if the first patent application, directed to the initial discovery, and the second patent application, directed to the improvement, do not name identical inventors, the first application may become available as prior art against the second application under certain circumstances. If not all inventors are employed by the same company the situation becomes more complex. If the initial discovery and the improvement are not commonly assigned, it is possible that the initial discovery may qualify as prior art against the improvement even if the initial discovery is never published and is never the subject of a patent application.

            These considerations are discussed below together with recommendations for increasing the likelihood of obtaining patent protection for both the initial discovery and the improvement. The first hypothetical involves the situation where the original research team collaborates with additional researchers at the same company. In the second hypothetical, the collaboration involves researchers at different institutions.

II.      Collaborations within a Company

            Assume that individuals α and β, employed at Company, discover that a novel antigen, Antigen A, provides a low level of immunity in mice against a pathogen. Company files a patent application naming α and β as co-inventors and claims Antigen A and its use as a vaccine against the pathogen. α and β then approach γ, also employed by Company, who specializes in adjuvants. Based on her expertise, γ selects a number of adjuvants for testing. After 10 months of testing, one adjuvant, Adjuvant X, is shown to reproducibly boost the immune response to Antigen A to the extent that nearly all immunized mice are protected from pathogenic challenge. Company files a second application a year after the filing date of the first application claiming Antigen A/Adjuvant X and its use as a vaccine. Company plans to market the improvement Antigen A/Adjuvant X as a vaccine for humans.

            A first issue that must be considered is the inventorship of the second patent application. If the selection of Adjuvant X would have been obvious to one of ordinary skill in the art at the time the improvement was made, perhaps it could be argued that γ is not an inventor of the second patent application since γ contributed nothing novel and nonobvious to the invention. However, both Federal Circuit precedent and practical considerations suggest that inventorship determinations cannot be based solely on this type of an analysis. First, in Burroughs Welcome Co. v. Barr Laboratories Inc. the Federal Circuit considered whether a United States Patent and Trademark Office (PTO) determination that certain claims were obvious over other prior art claims was relevant to a determination of whether the inventors of the two sets of claims were the same. In this case, the obvious claims were directed to a method of increasing the number of T lymphocytes with AZT. The prior art claims were directed to a method of treating AIDS with AZT. In holding that the obviousness determination was not controlling, the court concluded, "[f]or conception, we look not to whether one skilled in the art could have thought of the invention, but whether the alleged inventors had in their minds the required definite and permanent idea." Second, nonobviousness of an invention can be established long after a patent issues upon consideration of the so-called "secondary factors" of patentability. Thus, if inventorship were determined based on an obviousness analysis, inventorship would theoretically never be settled during the life of a patent.

            For the purpose of this hypothetical, we will assume that neither α nor β conceived of Adjuvant X. Therefore, regardless of whether the incorporation of Adjuvant X would have been obvious in light of the initial discovery, the holding in Burroughs Welcome suggests that γ should be named as an inventor of the second patent application.

            When it issues as a patent, the disclosure of the first application becomes prior art under 35 U.S.C. § 102(e) as of the date it was filed at the PTO. Since the invention claimed in the second application was invented "by another," the improvement may be held to be obvious over the disclosure of the first application by the PTO or a court even though there is common ownership. Thus, before permitting the first application to issue, Company’s patent attorney should weigh the probable strength of an obviousness rejection based on the first application against a claim specifically directed to the improvement. If there is a reasonable likelihood that such an obviousness rejection would be difficult to overcome, Company should consider incorporating the text of both applications into a continuation-in-part (CIP) application that claims both the initial discovery and the improvement and names α, β and γ as co-inventors. After the CIP has been filed claiming priority under 35 U.S.C. §120 to the first and second applications, the first and second applications should be abandoned. Hence, the initial discovery and the improvement would be prosecuted out of the CIP and the potential §102(e) prior art problem never materializes because no patent "by another" could issue due to the abandonment of the first application.

            As a caveat, in the above hypothetical, assume that the first application was filed one year prior to the filing date of the second application. These filing dates would not necessarily have affected the patent term if the filings occurred prior to June 8, 1995, as such patent terms end 17 years from the date of issuance. However, on June 8, 1995, the General Agreement on Tariffs and Trade Act (GATT Act) took effect, which implemented Article 33 of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs). Under the GATT Act, the patent term still begins upon the date a patent issues, but it now expires 20 years from the earliest U.S. filing date. Thus, by filing a CIP and claiming priority to both the first and second applications, the U.S. filing date for the purpose of calculating the patent term becomes the filing date of the first application, which would cause a one year loss of patent term for claims specifically directed to the improvement. If Antigen A/Adjuvant X turns out to be a multimillion dollar a year therapy, a one year loss in patent term could be worth millions of dollars. On the other hand, this monetary loss is perhaps mitigated by the following consideration: by combining the first and second applications into a CIP, Company is able to obtain claims specifically directed to the improvement and also generic claims that would potentially dominate Antigen A in combination with any other compound, which makes it difficult for a competitor to invent around the commercial embodiment simply by using a different adjuvant.

            The above strategy for overcoming Company’s own § 102(e) prior art was made possible by the enactment of the Patent Law Amendments Act of 1984, which amended the first paragraph of 35 U.S.C. §116 to read:

When an invention is made by two or more persons jointly, they shall apply for patent jointly and each make the required oath, except as otherwise provided in this title. Inventors may apply for a patent jointly even though (1) they did not physically work together or at the same time, (2) each did not make the same type or amount of contribution, or (3) each did not make a contribution to the subject matter of every claim of the patent.

Thus, in the above hypothetical, it is possible to prosecute the initial discovery and the improvement in the same CIP application even though inventor γ did not make a conceptual contribution to every claim of the CIP.

            Avoiding a § 102(e) rejection by combining two commonly owned applications into one CIP was specifically approved by the PTO in published explanatory comments when it promulgated rules under amended § 116. However, as a caveat, at least one commentator has urged caution before following this strategy in light of the traditional collaboration requirement of joint inventorship under certain circumstances. Citing Donald Chisum, other statutes and the legislative history of the Patent Amendments Act of 1984, this commentator concludes that "Congress's relaxation of the strict 'all claims' rule does not show an intent to abrogate the collaboration, or joint manner, requirement." Indeed, retaining the collaboration requirement for joint inventorship would appear to be supported by Federal Circuit case law since 1984.

            Prior to 1984, at least in certain circuits, all the inventors named in an application were also inventors of every claim in the application. Thus, the named inventors must have collaborated on every claim of the application. Following the 1984 amendment, the inventors of a particular claim in an application can differ from the named inventors of the application. While it is clear that the collaboration requirement survives the 1984 amendment of section 116, it is not clear whether compliance with the collaboration requirement is determined for the application as a whole or on a claim-by-claim basis.

            If the collaboration requirement is determined for the application as a whole, there must be some minimal collaboration between all named inventors even if different inventors contributed to separate, perhaps patentably distinct, claims. However, if compliance with the collaboration requirement is determined on a claim-by-claim basis, it would only be necessary that joint inventors of a particular claim collaborate, but not that all inventors named on the application collaborate.

            In our hypothetical, the collaboration requirement would be satisfied under either the application as a whole or the claim-by-claim analysis. Under the application as a whole analysis, the issue would be whether all the named inventors of the application at least minimally collaborated with each other in some way. For example, in the above hypothetical, assume that claim 1 in the CIP is directed to the Antigen A compound and claim 2 is directed to the Antigen A/Adjuvant X composition. Because α, β, and γ all collaborated in the joint invention of claim 2, under the application as a whole analysis, the collaboration requirement is satisfied. Under the claim-by-claim analysis, one looks to the inventorship of each claim, including assessing compliance with the collaboration requirement separately for each claim. Once inventorship is determined for each claim, the inventors named on the application are merely the summation of the inventors named for each claim. In the hypothetical, since α and β collaborated during the conception of the invention defined by claim 1, and α, β, and γ collaborated during the conception of the invention defined by claim 2, the collaboration requirement would also be met under the claim-by-claim analysis.

            However, this would change if our hypothetical is modified so that a claim 3 is added directed to Antigen B, which was solely and independently conceived by δ. If we assume that δ did not collaborate with α, β, or γ during the conception of Antigen B, the collaboration requirement would presumably not be satisfied based on the application as a whole analysis, but would be satisfied under a claim-by-claim approach.

            A claim-by-claim approach appears to have been favored by the PTO when it promulgated 37 C.F.R. § 1.45(c),

[i]f multiple inventors are named in a nonprovisional application, each named inventor must have made a contribution, individually or jointly, to the subject matter of at least one claim of the application and the application will be considered to be a joint application under 35 U.S.C. 116.

This suggests that the claim-by claim approach to the collaboration requirement is proper absent evidence to the contrary.

            In summary, a viable strategy for Company to overcome a potential obviousness rejection based on prior art under § 102(e) is to abandon the first and second applications after filing a CIP and prosecuting both the initial discovery and the improvement in one application. However, Company should keep in mind that naming multiple inventors in one application may only be proper if there was at least some minimum element of collaboration between each of the named individuals even though they may not have contributed conceptually to every claim.

III.    Collaborations with an Industrial Partner

            In the above hypothetical, assume that α and β are employed by Company, but γ is employed by Industrial Partner (IP). Thus, the initial discovery claimed in the first application (the Antigen A compound) would be assigned to Company. However, since γ is employed at IP, the improvement (the Antigen A/Adjuvant X composition) would be assigned to both Company and IP. Thus, not only would there be a potential obviousness rejection under §§102(e)/103 against a claim to the improvement in the second application if the first application were to issue as a patent, there would also potentially be an obviousness rejection under §§ 102(g)/103 based on the prior reduction to practice of Antigen A by α and β. Further, if communications between co-inventors during the development of an invention can constitute prior art, there would potentially also be an obviousness rejection under §§ 102(f)/103 based on α and β’s communication of Antigen A to γ. Moreover, the obviousness rejections based on §102(f) and §102(g) could not be overcome by combining the first two applications into a CIP. Sections 102(f) and 102(g) of the patent statute and their applicability to the present hypothetical are discussed in more detail below.

            A.        Are Sections 102(f) and 102(g) Prior Art Sections?

            Whether § 102(f) is indeed a prior art section, making 102(f) art available in obviousness determinations, has been the subject of much controversy. The literal wording of §102(f) provides that a person shall be entitled to a patent unless "he did not himself invent the subject matter sought to be patented." This has traditionally been considered a derivation provision, the purpose of which is to prevent someone from obtaining a patent on that which was invented by someone else. It does not only pertain to public knowledge, but also applies to private communications, even those made under a secrecy agreement.

            According to case law, two elements are required for establishing derivation under § 102(f). First, "the named inventor in the patent [must have] acquired knowledge of the claimed invention from another." Second, there must have been a prior conception of the invention, and the conception must have been communicated. In patent law, an invention is deemed to have been conceived when there is a "formation in the mind of the inventor, of a definite and permanent idea of the complete and operative invention." Thus, a prior reduction to practice of the invention by another is not required to establish derivation under § 102(f), provided that there was a prior conception and a communication.

            However, the case law has been in a state of disarray concerning whether a derivation under § 102(f) can be combined with other prior art in an obviousness rejection under § 103. For example, the Federal Circuit stated in dictum in New England Braiding that "[t]o invalidate a patent for derivation of invention, a party must demonstrate that the named inventor in the patent acquired knowledge of the claimed invention from another, or at least so much of the claimed invention as would have made it obvious to one of ordinary skill in the art." This dictum in New England Braiding conflicts with dictum in In re Bass and also conflicts with the dissenting opinion in Lamb-Weston, Inc. v. McCain Foods, Ltd. .

            Further confusion appears in Gambro Lundia AB v. Baxter Healthcare Corp. In Gambro, while referring to the dictum in New England Braiding, the lower court concluded that to invalidate Gambro’s patent, "Baxter did not need to prove communication of the entire conception, but rather only so much of the invention 'as would have made it obvious to one of ordinary skill in the art.'" However, the Federal Circuit held that, its dictum in New England Braiding notwithstanding, the lower court’s holding was clearly erroneous and "applied the wrong legal standard" because it "introduces incorrectly an obviousness analysis into the test for derivation." According to the Federal Circuit, the proper standard for finding communication of a prior conception was enunciated by the Supreme Court over 125 years ago in Agawam Woolen Co. v. Jordan, wherein the Court required a showing that the communication "‘enabled an ordinary mechanic, without the exercise of any ingenuity and special skill on his part, to construct and put the improvement in successful operation.’" Further, the Federal Circuit pointed out that this enablement standard for finding communication of a prior conception had been consistently applied by the Federal Circuit’s predecessor, the CCPA.

            The panel in Gambro, which was decided in April of 1997, included Circuit Judge Rader, Chief Judge Archer and Circuit Judge Lourie. Later, in August of 1997, a different panel of the Federal Circuit (Circuit Judges Michel, Lourie, and Rader) again turned the derivation standard on its head in OddzOn Products, Inc. v. Just Toys, Inc. While citing §103(c) of the patent statute, the OddzOn panel stated that

[w]hile the statute [§ 103(c)] does not expressly state in so many words that § 102 (f) creates a type of prior art for purposes of §103, nonetheless that conclusion is inescapable; the language that states that § 102(f) subject matter is not prior art under limited circumstances clearly implies that it is prior art otherwise. That is what Congress wrote into law in 1984 and that is the way we must read the statute.

Interestingly, the Federal Circuit in OddzOn did not cite Gambro, which was decided four months earlier, even though Gambro enunciated a different legal standard for derivation. According to the Federal Circuit in Newell Co., Inc. v. Kenney Manufacturing Co., where there is a conflict in statements of Federal Circuit law, the earlier statement prevails unless or until it has been overruled en banc. Thus, if the issue of whether § 102(f) is a prior art section is again before the Federal Circuit, it would appear that a subsequent panel (if it followed Newell) would presumably be forced to hold in the negative and follow the enablement standard as enunciated by the panel in Gambro. In our view, an en banc decision will be necessary to fully resolve this issue.

            In contrast to § 102(f), the case law is rather well settled that art qualifying under § 102(g) is available as prior art in an obviousness analysis, subject to the limitations discussed in subsections B and C below. Under the first sentence of § 102(g), a person is not entitled to a patent if "before the applicant’s invention thereof the invention was made in this country by another who had not abandoned, suppressed, or concealed it." While § 102(g) is typically associated with interferences, it may also be a grounds for invalidity in other contexts such as a defense in patent infringement litigation and during ex parte prosecution. In order to constitute prior art under §102(g), the invention must have been reduced to practice in the United States and it cannot have been abandoned, suppressed or concealed. The reduction to practice can be an actual reduction to practice, by actually completing the invention, or a constructive reduction to practice, by filing a patent application that satisfies the requirements of 35 U.S.C. § 112, first paragraph.

            A reduction to practice (actual or constructive) does not constitute prior art under §102(g) if the invention was abandoned, suppressed, or concealed. In order to constitute an actual reduction to practice the invention must be reduced to a physical or tangible form. With few exceptions, the invention must have been tested to confirm that it works for its intended purpose. Moreover, §102(g) does not contain a "personal knowledge requirement" or a "known to the art requirement." Thus, provided that there has been no abandonment, suppression, or concealment, under § 102(g) the prior work of another can constitute secret prior art against the work of a second inventive entity even where the prior work was not publicly known in the art and second inventive entity had no knowledge of it.

            B.        The § 103(c) Exception to 35 U.S.C. §§ 102(f) and 102(g) as Prior Art Sections.

            35 U.S.C. §103 states in part that "[a] patent may not be obtained [even though the claimed invention may be novel] . . . if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art . . . ." Thus, a claimed invention must not only be novel, but also nonobvious to be patentable. Importantly, a combination of prior art references or prior art "events" (e.g., a prior reduction to practice by another) can be used to formulate an obviousness rejection under § 103.

            In In re Bass , the PTO rejected claims in a patent application naming Bass, Jenkins, and Harvat as co-inventors as being obvious over previously filed issued patents to Bass and Jenkins. The PTO’s rationale was that the claims in the later filed patent application were merely an obvious variation of what had been reduced to practice in the earlier filed issued patents and thus the patent application was properly rejected under § 103 by virtue of § 102(g). Importantly, one of the two earlier filed issued patents was ultimately deemed to be by "another" for purposes of the § 102(g) analysis even though there was an overlapping inventor. In other words, the rejection was deemed to be applicable because the "inventive entity" of the earlier filed issued patents was not identical to the "inventive entity" of the later filed patent application. Also, the rejection was deemed appropriate even though the invention and the earlier reduction to practice occurred within the same company. The PTO’s rejection was upheld by the Court of Customs and Patent Appeals (CCPA), which was the predecessor court of the Court of Appeals for the Federal Circuit (Federal Circuit).

            In 1984, Congress amended § 103 to disqualify events that fall exclusively within §§ 102(f) or (g) from use as prior art under § 103, if specific conditions are met. This amendment can now be found in subsection (c) of § 103, which provides in pertinent part:

Subject matter developed by another person, which qualifies as prior art only under subsection (f) or (g) of section 102 of this title, shall not preclude patentability under this section where the subject matter and the claimed invention were, at the time the invention was made, owned by the same person or subject to an obligation of assignment with the same person.

13 U.S.C. §103 (c).

            This amendment was reportedly in direct response to the Bass decision, which arguably created a disincentive to file patents early and discouraged communication between co-workers. In support of the amendment, commentators argued that "'[s]uch encouragement of ignorance defeats a fundamental principle of corporate research - the free exchange of ideas between corporate employees. Moreover, it runs counter to both the policy and the spirit of the patent laws because it discourages both invention and the prompt disclosure of new inventions.'" Thus, recognizing the value of team research within corporations, businesses, and universities, Congress amended §103 to eliminate these obstacles to team research.

            Interestingly, the enacted version of the 1984 amendment to § 103 is considerably different from the version first proposed by Representative Kastenmeir in the House of Representatives on November 18, 1983. The original bill, entitled H.R. 4525, provided that "[p]rior art shall not include unpublished information which is developed by the applicant singly or jointly with others, or which is known to the applicant only by virtue of his or her employment." Thus, the original version of the amendment disqualified two types of art:

1) that "developed by the applicant singly or jointly with others," and

2) that "known to the applicant only by virtue of his or her employment."

Apparently, both types of disqualified prior art were to be limited to unpublished information.

            However, the original version was objected to both by the American Intellectual Property Law Association (AIPLA) and the PTO as being vague and overly broad. In particular, the AIPLA proposed redrafting the original version of the bill because:

We believe that [the original draft] in addition to modifying subparagraphs (f) and (g) of Section 102 of Title 35 might be interpreted as eliminating other prior art bars. Also, some of the words in [the original draft] are unnecessarily vague. The purpose of the amendment is to precisely define the needed remedy.

The PTO’s reason for supporting a revision was as follows:

Drafting an appropriate provision . . . has proven to be an elusive and complex task [The original draft of the amendment to section 103] is too broad. It is not limited, for example, to exchanges of background information among co-workers in a single organization. [Under the original draft] [i]nformation learned from or transmitted to outsiders could be disqualified as prior art.

Congress’ own explanation of the differences between the original and enacted versions of the amendment to §103 was that "the language in [the adopted version] is parallel to but also is more precise than the language of [the originally proposed version]. For example, [the enacted version] makes clearer that information learned from or transmitted to persons outside the inventor’s immediate organization is not disqualified as prior art."

            Thus, the enacted version of the amendment to § 103 (now § 103(c)) disqualifies "subject matter developed by another person" if it meets the following criteria:

            1) it falls only within the definitions of sections 102(g) or (f); and

            2) the subject matter of the prior art and of the claimed invention were:

                        a) commonly owned,

                        b) "at the time the [second] invention was made."

            When the PTO implemented § 103(c) by amending 37 C.F.R. §1.104, it explained that common ownership by the same "person" or "organization" "would include circumstances where the ownership resided in more than one person and/or organization as long as the applications are owned jointly by the same owners." In other words, for the § 103(c) exception to apply, there must be an identical "ownership entity" between the subject matter which would otherwise qualify as prior art under §§ 102(f) or (g) and the subsequent invention claimed in the later filed patent application. In other words, to receive benefit of § 103 (c), if the prior art subject matter was owned by companies A and B, then the subsequent invention must also have been owned by companies A and B at the time the subsequent invention was made.

            This interpretation by the PTO of "commonly owned" appears to be consistent with Congress when it amended the original bill in order to" make clearer that information learned from or transmitted to persons outside the inventor’s immediate organization is not disqualified as prior art."

            C.        An Inventor’s Own Prior Work as Prior Art Under 35 U.S.C. §§ 102(f) and 102(g).

            Assuming that the holding in OddzOn is adopted by the Federal Circuit generally, it means that

an invention, A', that is obvious in view of subject matter A, derived from another, is also unpatentable. The obvious invention, A', may not be unpatentable to a third party who did not receive the disclosure of A, but it is unpatentable to the party who did receive the disclosure.

The standard enunciated in OddzOn is perhaps workable where the subject matter is derived from a non-inventor. However, it becomes problematic where the subject matter is derived from a co-inventor. For example, in our hypothetical, α, β and γ are co-inventors of the improvement (i.e., the Antigen A/ Adjuvant X composition), which is potentially obvious over the initial discovery of Antigen A by α and β. Since the second inventive entity (α, β and γ) is considered to be "different" than the first inventive entity (α and β), the second inventive entity could be said to have derived subject matter (the initial discovery Antigen A) from the first inventive entity that renders the improvement obvious when combined with other prior art documents.

            Thus, an important unanswered question is whether the Federal Circuit's holding in OddzOn would apply to communications between inventors. In 1982, the Fifth Circuit in Shields v. Halliburton Co. held that the prior work of one inventor does not constitute prior art against a later joint invention between the inventor and another. In reaching this conclusion, the Fifth Circuit relied heavily on the fact that no patent had been filed directed to the first inventor's initial discovery. While Chisum criticizes the rationale behind the Fifth Circuit's holding, Chisum agrees that communications between co-inventors should not constitute prior art under § 102(f).

In at least one instance derived knowledge should not be treated as "prior art." This is the case of joint invention. The prior knowledge in fact developed by a joint inventor as part of a prior invention should not be "prior art" under Section 102(f) as to the joint invention. This would be an exception to the normal concept of separate "inventive entities." Thus, the prior secret work of one joint inventor would be prior art as to the later invention of joint inventors only under Section 102(g)—which assumes reduction to practice and no abandonment, suppression or concealment.

Arguably, applying the OddzOn standard to discussions between co-inventors at collaborating biotech institutions would serve to frustrate, rather than promote, discoveries in biotechnology. This is particularly true since, in order to be joint inventors, there must have been some degree of collaboration among the inventors. As the Fifth Circuit said in Shields, "if the first inventor’s initial work . . . constitutes an earlier invention to any subsequent effort with a collaborator, no valid joint invention would be possible. Theoretically every joint invention would have to be the result of simultaneous inspiration of the collaborators." However, like Chisum's admonishment of the Fifth Circuit's rationale in Shields on the grounds that § 102 "provides no basis for distinguishing between patented and unpatented prior invention insofar as the definition of 'another' is concerned," the patent statutes provide no basis for making a distinction between prior art under §102(f) versus prior art under §102(g) insofar as the prior work of one of the inventors is concerned. In other words, we fail to see a statutory basis for the Federal Circuit to "carve out" an exception where an inventor’s prior work qualifies as prior art under § 102(f), but not recognize a similar exception where the inventor’s prior work qualifies as prior art under § 102(g).

            In our hypothetical, since Company and Company-IP are different ownership entities and thus § 103(c) would not apply, obviousness issues could be presented under §§ 102(g)/103 against the improvement invented by α, β and γ based on the prior reduction to practice of Antigen A by α and β. Thus, even if the Federal Circuit were to adopt Chisum’s recommendation and recognize the exception that "knowledge developed by a joint inventor as part of a prior invention should not be 'prior art' under Section 102(f) as to the joint invention," patenting improvements arising out of collaborations between companies would still prove problematic unless this exception is extended to reach a joint inventor’s previous reduction to practice that qualifies as prior art against the joint invention under § 102(g). However, if applied generally, the holding in In re Bass, discussed supra, would appear to militate against recognizing such an exception to prior art under § 102(g) . Thus, another important question is whether the holding in In re Bass creates a per se rule of unpatentability under §§ 102(g)/103 based on a joint inventor’s previous reduction to practice.

            In General Motors Corp. v. Toyota Motor Co., Ltd. , decided prior to the enactment of § 103(c), at issue before the Sixth Circuit was the obviousness of a catalytic converter that had been developed at General Motors (GM) in at least three stages. While citing In re Land and In re Bass, Toyota argued that the "three steps are discrete inventions because the first two steps did not result from the collaboration of the patentees" and thus the first two steps constituted prior art against the patented catalytic converter. GM countered with the argument that "there was only one invention, the patented converter, and the two earlier steps in its development should be seen as merging into the final product." The Sixth Circuit sided with GM, and "explained around" the CCPA’s holdings in In re Land and in In re Bass:

Neither Land nor Bass indicates that the prior inventions were in any way the product of concerted effort within a business entity. Under the facts of this case, where numerous "inventors" all worked under the aegis of one employer toward a common goal, it is appropriate to define the concept of joint invention broadly. It is not realistic to require in such circumstances that joint inventors work side-by-side, and that each step in the inventive process be taken by all the firm’s collaborators.

If it is not realistic to require that each step in the inventive process be taken by all the collaborators while working at a single institution, it is arguably even less realistic to require this where the collaborators are employed at different institutions. However, it is presently unclear whether the Federal Circuit would follow the Sixth Circuit’s rationale in General Motors and extend it to situations involving collaborators working toward a common goal or goals while employed at different institutions. In our view, recognizing α and β’s initial discovery as prior art under §§102(g)/103 against α, β and γ’s improvement places an unreasonable burden on both Company and IP. It places Company in the unenviable position of having to decide whether to abandon, suppress, or conceal the initial discovery to better ensure that patent protection will be obtained for a potentially more commercially viable improvement and, for the same reason, gives the collaborating IP an incentive to seek such assurances from Company prior to entering into the collaboration agreement. Further, creating a legal setting whereby the abandonment, suppression, or concealment of an initial discovery can, in certain circumstances, be beneficial runs contrary to a stated purpose of the patent system, which is to encourage the early disclosure of inventions to the public.

            In sum, it is our view that the prior art effect of the previous knowledge or previous reduction to practice of an inventor against a later joint discovery by the inventor and another remains uncertain. Important unanswered questions include whether the Federal Circuit's holding in OddzOn (that prior art under § 102(f) can be relied on in an obviousness determination under

§ 103) would apply to communications between joint inventors and whether the CCPA’s holding in In re Bass creates a per se rule of unpatentability under §§ 102(g)/103 based on a joint inventor’s previous work. In light of the present day reality that collaborative research between biotech and pharmaceutical institutions is an integral part of the industry, the Federal Circuit may want to consider recognizing the equity on a case-by-case basis of defining joint invention broadly to encompass the previous knowledge and work of each joint inventor, which would arguably be in agreement with, or at least a logical extension of, the Fifth Circuit’s holding in Shields and the Sixth Circuit’s holding in General Motors.

            D.        Recommendations

            In our hypothetical, to avoid encountering potential obviousness rejections under § 102(f) and § 102(g), Company and IP should have at least considered an agreement whereby IP would contractually agree to assign its rights to all inventions arising out of the collaboration to Company. This would have invoked § 103(c) of the statute, which disqualifies §102(f) and §102(g) from use as prior art under § 103 if, at the time the improvement was made, the improvement and the initial discovery were "owned by the same person or subject to an obligation of assignment with the same person." If such a common assignment or an obligation to assign the improvement to Company had been in place, § 102(f) and § 102(g) prior art issues based on the initial discovery would be resolved. (The Company patent attorney could then combine the first and second applications into a CIP to prosecute claims to both the initial discovery and the improvement in one application as we discuss in Section II above to avoid any § 102(e) issues.) Thus, from a patent law perspective, crafting collaboration agreements such that § 103(c) applies best ensures that claims are obtained to both the initial discovery and the improvement because, when prosecuting the improvement, it removes the obstacle of having to overcome an obviousness rejection on the merits based on the initial discovery.

            However, from a business standpoint, agreeing to assign all inventions arising out of the collaboration to Company may not be acceptable to IP. This is because each joint owner (also called a co-owner or a tenant-in-common) is an owner of an undivided one-half interest in the patent and is said to be at the mercy of the other joint owners. The incidents of joint ownership are codified in 35 U.S.C. § 262 as follows:

In the absence of any agreement to the contrary, each of the joint owners of a patent may make, use, offer to sell, or sell the patented invention within the United States, or import the patented invention into the United States without the consent of and without accounting to the other owners.

Additionally, all co-owners must join in a patent infringement suit. Further, a "primary interest" of a co-owner is "the interest . . . in being able to license third parties under his or her patent without harassing suits by other co-owners." Thus, clear advantages flow from being a joint owner of a patent, which may make IP reluctant to enter into an agreement assigning its rights to inventions arising out of the collaboration to Company. However, this notwithstanding, many of IP’s concerns could presumably be addressed contractually, such as, for example, with an exclusive licensing arrangement.

            Alternatively, to avoid having to fashion a licensing arrangement that compensates for IP’s lack of ownership status, Company and IP could have agreed to assign Company’s initial discovery and all inventions arising out of the collaborative research to a joint venture, incorporated by Company and IP especially for the purpose of the collaboration. Another option would be for Company to agree, prior to the start of the collaboration, to assign to IP an undivided, one-half interest in any patent application directed to Company's initial discovery, with the assignment conditioned upon the development of a patentable improvements from the collaboration. Thus, at the time of the invention of Antigen A/Adjuvant X, because of Company's obligation to assign rights in Antigen A, and due to the collaboration in the development of Antigen A/Adjuvant X, both inventions would be under an obligation to be assigned to the same ownership entity: Company-IP. Therefore, under either alternative, at the time of the invention both the potential "prior art" under § 102(f) or (g)/§ 103 of Antigen A, and the improvement invention of Antigen A/Adjuvant X would be under an obligation of assignment to the same ownership entity, and would fall within the exception in § 103(c). However, one caveat of both alternatives is that Company may not be willing to share its rights in Antigen A with IP for only the possibility of the development of valuable improvements from the collaboration. Thus, none of the alternatives we have discussed completely abrogate the § 102(f) or (g)/§ 103 prior art problem that plagues collaborations between companies.

IV. Conclusions

            If a first patent application, directed to an initial discovery, and a later filed second patent application, directed to an improvement, do not name identical inventors, the disclosure of the first application may become available as prior art upon issuance into a patent against the second application under § 102(e). Thus, before permitting the first application to issue, the possibility of incorporating the text of both applications into a CIP that claims the initial discovery and the improvement should be considered. As a caveat, caution should be exercised before following this strategy in light of the traditional collaboration requirement for joint inventorship.

            If the second application is assigned to a different ownership entity than the first application, §§102(e)/103 may apply as above against a claim to the improvement in the second application once the first application issues as a patent. In addition, §§ 102(g)/103 may also apply based on the prior reduction to practice of the initial discovery. This is because § 103(c) does not apply where the initial discovery and the later discovered improvement are not commonly owned, or at least under an obligation of common assignment, at the time the improvement is made. Further, if communications between co-inventors during the development of an invention can constitute prior art, §§ 102(f)/103 may also apply. While the law remains unsettled regarding the propriety of obviousness rejections under § 102(f) or §102(g) based on a joint inventor’s previous knowledge or work, we agree with others in recommending that the Federal Circuit recognize that prior knowledge developed by one inventor should not constitute prior art under §102(f) as to later joint discoveries by the inventor and another. However, in our view, recognizing such an exception for § 102(f) but not for § 102(g) could promote the abandonment, suppression, or concealment of initial discoveries to avoid them from becoming secret prior art as to later discovered improvements arising out of the collaboration.

            Strategies for ensuring that § 103(c) applies to collaborative research include having the industrial partner agree to assign all rights to inventions arising out of the collaboration to the company making the initial discovery. Alternatively, the company and the industrial partner could agree to assign the initial discovery and all inventions arising out of the collaboration to a joint venture, formed especially for purposes of the collaboration, or the company could agree to conditionally assign an undivided, one-half interest in its first application to the industrial partner.