REGENTS OF UNIVERSITY OF MICH. & REPLIGEN CORP. v. Bristol-Myers Squibb Co.

301 F.Supp.2d 633 (2003)

The REGENTS OF THE UNIVERSITY OF MICHIGAN and REPLIGEN CORP., Plaintiffs,
v.
BRISTOL-MYERS SQUIBB CO., Defendant.

No. 00-CV-73690.

United States District Court, E.D. Michigan, Southern Division.

September 10, 2003.

J. Michael Huget, John C. Blattner, Ann Arbor, MI, for Plaintiffs.

Edward H. Pappas, Dickinson Wright, Bloomfield Hills, MI, Thomas J. O'Connell, Washington, DC, for Defendants.

FINDINGS OF FACT AND CONCLUSIONS OF LAW

STEEH, District Judge.

A bench trial was held in the above-captioned case beginning on April 2, 2003 and concluding on May 5, 2003. The Court commends counsel on this case for doing an extraordinary job of making the subject matter easy for the court to understand. Reading the voluminous filings in this case, and presiding over the fourteen days of trial, while time-consuming, was not unenjoyable, due to the quality and professionalism of the lawyers and everybody else involved in the case. Both sides have submitted proposed findings of facts and conclusions of law, which the Court has thoroughly considered. Now therefore, for the reasons stated below, judgment in this case shall enter for defendant.

FINDINGS OF FACT

This is an inventorship case wherein plaintiffs, the Regents of the University of Michigan and Repligen Corporation, seek to have Dr. Craig Thompson named a co-inventor on certain patents pursuant to 35 U.S.C. § 256. The patents-in-suit, which are held by defendant Bristol-Myers Squibb Company, relate to a protein which can be found on T-cells, and which can be used to regulate the immune response.

A number of individuals play a role in determining the issue of inventorship. Craig B. Thompson, M.D., was, during the time period relevant to this inventorship question, an employee of the Howard Hughes Medical Institute and an Associate Professor at the University of Michigan. Peter S. Linsley, Ph.D., was a senior research scientist with Bristol's Oncogen Division in Seattle, Washington. Jeffrey A. Ledbetter, Ph.D. and Nitin K. Damle, Ph.D. were also senior research scientists with Oncogen. William Brady was a research scientist with Oncogen. Carl June, M.D., was an active-duty medical officer in the United States Navy at the Naval Medical Research Institute.

The two expert witnesses were of great assistance and provided the Court with at least a basic understanding of the science involved in this case. Plaintiffs' expert witness was Dr. Lawrence Stern, an Associate Professor at the University of Massachusetts Medical School. Defendant's expert was Dr. Jeffrey Ravetch, Professor and head of the Laboratory of Microbiology and Immunology at the Rockefeller University.

An immune response requires B-cell and T-cell interactions, called recognition and co-stimulation. It is now understood that co-stimulation is necessary for an immune response and involves the binding of the B7 protein on B-cells with the CD28 protein on T-cells. The binding between B7 and CD28 was discovered by Dr. Peter Linsley, Ph.D., a Bristol scientist, on December 2, 1989. Blocking the co-stimulatory binding of CD28 to B7 is a means to shut down or prevent an immune response. In order to block the binding of CD28 to B7, one must have a molecule that will also bind to B7, but with a higher affinity, thereby preventing the CD28/B7 interaction.

CTLA4 is a protein found on the surface of activated or cytotoxic T-cells. CTLA4 is a member of the immunoglobulin ("lg") superfamily of genes, as is CD28. CTLA4 is activated when its ligand, also B7, binds to it. The activation of CTLA4 results in an inhibitory signal, turning off the immune response, being sent by CD28 into the T-cell. CTLA4 is not present on the surface of resting or non-activated T-cells. Following activation of the T-cell, CTLA4 is expressed on the surface of the cell. This expression of CTLA4 on the surface of the T-cell serves to down regulate T-cell activity by (a) competing with CD28 for B7, thereby reducing the CD28 pathway activation, and (b) when B7 binds to CTLA4 sending a signal into the cell to turn off the activation of the T-cell.

In general, CD28 and CTLA4 serve as on and off signals for the T-cell, activating the T-cell and consequent immune response and subsequently turning off the immune response. While the signals sent by CD28 and CTLA4 are read differently by the cell, the extracellular domains perform the same function, i.e., they both are activated by binding B7. Activation of the immune response is important in the treatment of diseases where heightened immune function is desirable, for example therapy of cancer or infection. Inhibition of the immune response is of great interest in the treatment of auto-immune diseases, where the immune system is attacking the host organism and also in organ transplantation.

It is now known that CD28 and CTLA4 perform opposing functions in the immune system. When CD28 binds to B7, that stimulates the T-cell and activates the immune response. It is the "on" switch. When CTLA4 binds B7, it deactivates the T-cell and shuts down the immune response. It is the "off" switch.

CD28 and CTLA4 are both members of the single V domain family of the lg superfamily, which consists of approximately 400 members. The function of a large number of these molecules, including CTLA4, was unknown in 1989-1990.

The patents-in-suit relate to soluble derivatives of CTLA4, such as CTLA4-lg, which can be used to block co-stimulation and thereby regulate the immune response, Bristol is the assignee of all rights in the inventions claimed in the seven patents-in-suit:

U.S. Patent No. 5,434,131 ("the '131 patent"),

U.S. Patent No. 5,844,095 ("the '095 patent"),

U.S. Patent No. 5,851,795 ("the '795 patent"),

U.S. Patent No. 5,885,579 ("the '579 patent"),

U.S. Patent No. 5,968,510 ("the '510 patent"),

U.S. Patent No. 5,977,318 ("the '318 patent"), and

U.S. Patent No. 5,885,796 ("the '796 patent").

Each of these patents names Drs. Peter Linsley, Jeffrey Ledbetter, Nitin Damle, and Mr. William Brady as inventors. Four of the patents also name Dr. Peter A. Kiener as an inventor. Four of the patents, namely the '095, '795, ' 796, and '318 patents cover CTLA4 fusion proteins that bind B7, soluble and full length CTLA4 proteins and nucleic acid molecules encoding the same, methods for producing soluble CTLA4 proteins, as well as plasmids, host vector systems and monoclonal antibodies using CTLA4. The remaining three patents, the '131, '579, and '510 patents, cover methods of using soluble CTLA4 protein products, such as fusion proteins, to regulate the immune system by affecting B7 binding.

Beginning of the Collaboration

Plaintiffs describe a scenario of an extensive collaborative effort undertaken by the named inventors and Thompson, as reflected in a joint patent application, numerous co-authored articles, many telephone conversations, meetings and the sharing of reagents and prepublication experimental results, plans and concepts. According to plaintiffs, the collaboration included more than the activation of the CD28 pathway; it included the role of cell surface receptors and the T-cell generally and the activation and inhibition of the CD28 pathway in particular.

Defendant attacks plaintiffs' story by pointing out that general references to unspecified conversations, meetings, reagents, plans, and concepts do not establish a contribution to the specific inventions claimed in patents-in-suit. Defendant maintains that there is no clear, convincing and corroborated record evidence that Thompson made an inventive contribution to any claim of the Bristol patents. Rather, the record evidence supports the conclusion that CTLA4 was not part of the collaboration with Bristol and Bristol independently designed, made and tested CTLA4-lg.

In 1986, Drs. Ledbetter, Carl June, and Thompson began their collaboration by investigating the role of "accessory" cell surface molecules on the T-cell, i.e., cell surface molecules other than the T-cell receptor ("TCR"). The objective of the collaboration was to articulate the basic science underlying these accessory molecules which appeared to play a role in activating T-cells. During the period from 1986 to 1990, Thompson, June and Ledbetter collaborated on and studied a host of accessory molecules in addition to CD28, but not CTLA4. The collaboration came to focus on elucidating the details of what they discovered to be an important pathway of T-cell regulation that was mediated by the cell surface molecule CD28. They determined that a powerful and independent means of modifying the immune response might be possible through regulation of the CD28 pathway.

Over time, the collaboration grew to include Drs. Tullia Lindsten, M.D., Ph.D., Lawrence Turka, M.D., and Kelvin Lee, M.D., all from Thompson's laboratory, as well as Dr. Linsley, from Bristol. The collaborators published over a dozen collaborative articles. In November, 1988, Ledbetter, June, Thompson and Lindsten filed a joint patent application entitled "Immunotherapy Involving CD28 Stimulation."

In January 1989, Dr. Linsley began working with Dr. Ledbetter on experiments concerning the CD28 pathway. In the context of this research, Dr. Linsley entered the collaboration. During this same period, from 1988 to 1990, the collaborators also pursued research projects on the CD28 pathway independently of one another. Thompson focused on inhibition of the CD28 pathway in connection with the treatment of autoimmune diseases such as sepsis, trauma and arthritis in ongoing studies funded in 1988 and 1989 by the grant from the Navy and a similar grant from the University of Michigan. In these grants, Thompson stated that a soluble form of CD28 might block the activation of cell surface CD28 by binding with its then unknown ligand, thereby blocking the ligand's interaction with CD28, and thereby prove beneficial in treating these autoimmune diseases. In connection with this research, Dr. Thompson was investigating whether a natural ligand for the CD28 receptor existed. Dr. Linsley of Bristol was also investigating whether a natural ligand existed for CD28. It was Dr. Linsley who discovered that B7 was the natural ligand for CD28 on December 2, 1989.

According to Dr. Thompson, each of the collaborators spoke every couple of weeks, or more frequently as matters of interest arose. Thompson describes the collaborators as openly sharing concepts, reagents, research ideas and results. This description of the collaboration comes from Dr. Thompson and is corroborated by June and Turka.

Thompson's Role in Collaboration

In April 1989, Dr. Lindsten attended a seminar in Israel at which she learned of a recently identified gene named CTLA4, the sequence of which had been published in an article by the Golstein laboratory. The literature at this time described CTLA4 as being a member of the superfamily of lg genes. The literature also pointed out the structural similarity and close sequence homology between CTLA4 and CD28. Thompson informed Linsley of CTLA4. Prior to this disclosure, none of the collaborators were aware of CTLA4.

Around this same time, Kelvin Lee, a post-doctoral research scientist in Thompson's laboratory, was working on delineating the genomic structure of CD28. Using the genomic sequence of CTLA4, which had been published previously by others, and Lee's pre-publication genomic organization of CD28, Thompson hand aligned the extracellular domains of CTLA4 and CD28. Thompson found the homology to be more impressive in certain domains than in the overall homology of the two genes. Such homologous genomic structures provided a strong indication that the genes arose as a result of a gene duplication.

When a gene duplicates, at the moment of duplication, the protein products of the two genes bind the same ligand. Thompson described how he hypothesized that CD28 and CTLA4 might continue to bind the same, or a related, ligand, based on his comparison of the genomic organization of each protein. Dr. Thompson did not publish his gene duplication theory, nor did he record his findings in writing. During plaintiffs' rebuttal case. Dr. Thompson testified that he shared his gene duplication hypothesis with Dr. Linsley in late-April 1989. Dr. Thompson's live rebuttal testimony is inconsistent with his deposition testimony, offered during plaintiffs' case-in-chief, where he stated that he came up with the gene duplication hypothesis in December 1989, over the school vacation. Dr. Thompson also testified during his deposition that the earliest he remembered telling Dr. Linsley that CTLA4 might bind a B7 related molecule was late April 1990.

To support the scenario that Dr. Thompson shared his concept and hypothesis that CD28 and CTLA4 might bind the same ligand in late-April 1989, plaintiffs offered into evidence a letter dated May 15, 1989. The unsigned letter, which Dr. Thompson testified was sent to Dr. Linsley by overnight delivery, enclosed Lee's pre-publication genomic structure of CD28, which was virtually complete by this time. The letter invites Linsley to "feel free to use this information in organizing your own studies", and ends by saying "[i]t sounds like you have the first good evidence that CD28 has a natural ligand."

The May 15, 1989 letter does not corroborate Dr. Thompson's testimony that he told Linsley that CD28 and CTLA4 might bind the same ligand in late-April 1989. The letter does not refer to Thompson's alleged gene duplication theory. What is clear is that Thompson shared the Golstein article with Linsley in April 1989, and on May 16, 1989, Linsley downloaded a CTLA4 sequence from a public database. Linsley testified that he was prompted to download the CTLA4 sequence because of the fact that Golstein noted a structural similarity and close sequence homology between CTLA4 and CD28, not because of the May 15 letter.

Linsley was working on finding a natural ligand for CD28 during this period of the Spring of 1989. It was not until December 2, 1989 that Linsley would have the first transfection data showing that B7 was the natural ligand for CD28. The collaborators continued with their research, which was documented in several collaborative articles on the CD28 pathway. None of these articles mentioned CTLA4. Kelvin Lee completed his research on the genomic organization of CD28 in November 1989, though his laboratory notebook shows a last entry on January 10, 1990, and had begun drafting an article outlining his findings. Plaintiffs refer to the Lee research and article as corroboration for Thompson's gene duplication theory. While the CD28 genomic sequencing may well have been used by Thompson to come up with or confirm his theory, there is no corroboration that he communicated it to Linsley or anyone else at Bristol. Lee's article does not mention CTLA4, let alone compare it with CD28.^[1]

It is clear that Thompson became aware of Linsley's discovery that B7 was the natural ligand for CD28 in December 1989. Dr. Ledbetter wanted to put that discovery in a "note added in proof" to an article written by Lawrence Turka, a post-doctoral research scientist in Thompson's lab. The "Turka Article" was submitted to the Journal of Immunology for publication on December 7, 1989 and accepted on December 8, 1989.^[2] The Turka article related to the CD28 pathway, and the collaborators decided to report the CD28-B7 binding so that discovery would not be "scooped" by another laboratory.

Thompson testified that many discussions took place in mid-December 1989 regarding whether to include Linsley's CD28-B7 binding results as a "note added in proof" to the Turka Article. Thompson testified that during these discussions he again shared with Linsley his laboratory's pre-publication data on the genomic organization of CD28; his theory of gene duplication; his theory that CD28 and CTLA4 shared the same or a related ligand; and his suggestion that the next step for the collaborators was to test CTLA4 to determine if it bound to B7. Linsley denied taking part in the discussions described by Thompson, adding that he was unaware that there was a note added in proof until after the Turka Article was published.

The Review Article in Immunology Today was also drafted beginning in December 1989. Plaintiffs maintain that this review of the collaboration necessitated additional discussions among the collaborators. However, this article dealt only with CD28 and is irrelevant to the inventorship issue in this case.

Lee, Turka and June each testified that Thompson told them about his gene duplication theory toward the end of 1989. They pinpointed the time period based on Lee's sequencing of the genomic organization of CD28. This corroborates Thompson's deposition testimony that he came up with his gene duplication theory over the December holidays, and during the process of preparing Lee's article on the genomic organization of CD28. However, there is no corroboration for the proposition that Thompson shared his thoughts with Linsley, or any of the Bristol scientists, in December of 1989.

Ledbetter testified that he learned of Thompson's belief that CTLA4 might bind B7 from Linsley, and he believed Linsley learned it from Thompson sometime in 1989. However, Ledbetter was fuzzy on dates throughout his deposition, so his credibility as to this date is not high. Ledbetter testified that when he heard about Thompson's "speculation", he personally did not believe CTLA4 would bind B7. Linsley himself also testified that even if Thompson had communicated his hypothesis that CTLA4 would bind B7, which he did not recall ever happening, Linsley would not have given this unproven hypothesis much credence. At trial, Linsley explained that there was no way of knowing whether CTLA4 would bind B7 until the experiment was done, and he had already planned to do the experiment.

At trial, Linsley explained that he conceived that CTLA4 might bind B7 "within a day or a week" of his December 2, 1989 experiment confirming that CD28 binds B7. At this time, Linsley knew the nucleotide sequence of CTLA4, which was previously published, and had read articles that said that CTLA4 and CD28 shared significant structural homology. The 1988 Golstein article reported that CD28 and CTLA4 shared "some significant structural features, such as the precise number and relative position of cysteines" and recommending that "CTLA4 target adhesion" (binding) be investigated. Linsley never said he could predict CTLA4 would bind B7 because there was no scientific basis on which anyone could make that prediction. Linsley did testify it was conceivable that CTLA4 might bind B7, and he conceived of the experiment and reagents to test that. With all of the information available at that time, it was not illogical to test CTLA4 for its ability to bind B7. The fact that Linsley did not share his planned experiments with Thompson is evidence that he did not believe this line of experimentation was part of the CD28 pathway collaboration.

The collaborators set a meeting for July 23, 1990 to exchange information and ideas on their research to date and to plan the direction of their continuing work. Thompson testified that he reiterated his ideas that the collaborators should focus on making soluble forms of CD28 and CTLA4, and that, in particular, they should make CD28-lg and CTLA4-lg. Thompson suggested that Bristol use the lg fusion technology to make CTLA4-lg and then his laboratory would devise and perform the necessary biological tests. Dr. Ledbetter acknowledged that Dr. Thompson did suggest that the Bristol scientists make soluble CD28 and CTLA4, but that it was the first time the suggestion had been made to the collaborators by Dr. Thompson. Dr. Ledbetter testified that Dr. Thompson did not seem to be aware of the previous efforts within Bristol to express CD28-lg and CTLA4-lg fusion proteins. Dr. Ledbetter stated that after the meeting at which Dr. Thompson made his suggestion, he told Dr. Linsley that their plan to make CTLA4-lg should be moved to a high priority.

[Thompson] suggested that we take this approach. And it was a — was an approach that Peter [Linsley] and I were already working on. And the fact that Craig Thompson suggested that we make these molecules was — my response to it was that it wasa — was competitive in nature, and it suggested that we better move on doing this because Craig is now — has the same idea we do. We heard him say it, we were already working on it, but now Craig had the idea, as well. And so that encouraged us to move forward and move its priority to a high level.

Ledbetter deposition at pp. 191-92.

Before the meeting with Dr. Thompson, Dr. Linsley had done some planning to make the lg fusion protein. On March 23, 1990, Dr. Linsley designed primers to make CTLA4-lg. Linsley's lab began by making other lg fusion proteins, such as B7-lg. After having some success, on July 17, 1990, Mr. Brady commenced work on making CTLA4-lg by trying to identify a source of CTLA4. Brady was able to use some components from his work on other lg fusion proteins to expedite making CTLA4-lg. The comments by Dr. Thompson at the July 23, 1990 meeting did seem to cause Bristol to put more emphasis on the project, but do not suggest that they undertook something they were not otherwise preparing to do.

Immediately prior to the July 23, 1990 meeting, Linsley was attempting to map the location of CTLA4 and CD28 on the chromosome. Plaintiffs argue that Linsley engaged in this exercise as a result of Thompson's conception and in anticipation of the July 23, 1990 meeting. Linsley testified he was doing the gene mapping experiments to extend the research findings in a March 1990 article by Lafage et al.^[3] that reported that CD28 and CTLA4 were located on the same chromosome and could have been the result of a gene duplication event.

Bristol's Expression of CTLA4-lg

In late-March, Mr. Brady had made his first lg fusion construct, specifically B7-lg. Designing a fusion protein required fusing together specified pieces of DNA encoding three proteins. These three proteins are the signal sequence, the expression vector and the lg tail. By March 22, 1990, Dr. Linsley had designed an lg tail — lgG1 with mutations in the hinge region and Bcl I and Xba I sticky ends. This lgG1 tail would ultimately be used to make CTLA4-lg. By March 22, 1990, Bristol also had a CDM8 vector with Xba I and Hind III sticky ends which would ultimately be used to make CTLA4-lg. Finally, by March 22, 1990, Bristol had an OncoM signal sequence which would be used to make CTLA4-lg. Each of these components were unique and were not related to CTLA4 or lg in their natural states.

While at this time the general concepts for the construction of lg fusion proteins had been disclosed, it was not known how to make functional CD28-lg and CTLA4-lg fusion proteins. Another laboratory had made functional CD4-lg and B7-lg, but the published articles showed as many failures as successes. The techniques such as PCR and transfecting and growing cells were well established at the time. However, like building a skyscraper, building a fusion protein requires not only tools, but a design.

Dr. Linsley's design and strategy for making a CTLA4-lg fusion protein containing the extracellular portion of CTLA4, an OncoM signal sequence, an lg tail and a CDM8 vector, with appropriate sticky ends, was complete by March 23, 1990, as evidenced by his primer order form of that date. Plaintiffs question the authenticity of the March 23 primer order form because it was not pasted into a notebook and Bristol did not produce the 3-ring binders in which the primer order forms were stored. The March 23 primer order form was produced by Bristol from its business records, so the fact that it was not pasted in Linsley's notebook does not diminish its authenticity. As to the second criticism, the binders were discarded or lost when Bristol's Seattle facility was closed.

By June 29, 1990, Mr. Brady successfully made the CD28-lg fusion protein, also designed by Linsley, and provided it to Dr. Linsley. On June 29, 1990, Dr. Linsley tested the CD28-lg fusion protein and determined that it bound B7.

On July 17, 1990, Brady began his work on making a CTLA4-lg fusion protein. By July 18, Brady found a source for CTLA4 mRNA. Next, Brady split the 5' primer that Linsley designed on March 23, 1990 to make two shorter 5' primers. Mr. Brady's two shorter 5' primers were functionally the same as Linsley's longer 5' primer and would produce the identical amino acids. Brady's 5' primers contained the same OncoM leader sequence, sticky ends, junction between CTLA4 and lg, and overlap with CTLA4 as reflected in Dr. Linsley's March 23, 1990 primer order form. Brady's two shorter 5' primers differed from Linsley's March 23, 1990 longer 5' primer by one DNA base in one of the primers, but still encoded the same amino acid. Mr. Brady used a different base for the purely technical reason that it increased the yield and lowered the error rate during synthesis of the primes. Brady used the exact same 3' primer design which Linsley designed on March 23, 1990. On July 18, 1990, Mr. Brady ordered the two shorter 5' primers to be synthesized.

On July 26, 1990, Brady began his efforts to express CTLA4 by running the first PCR for its extracellular domain. On July 30, Brady created a DNA ring, complete with 5' primer, 3' primer, OncoM leader sequence, Hind III sticky end, lg tail and CDM8 expression vector. On August 1, 1990, Brady transfected the CTLA4-lg fusion plasmid into mammalian COS cells, which produced the CTLA4-lg fusion protein. On August 3, 1990, Brady collected the liquid medium containing the CTLA4-lg fusion protein and provided it to Dr. Linsley. While Mr. Brady's work to make CTLA4-lg took only about two weeks, the tasks of designing the primers and developing an expression system had already been worked out in the preceding months.

On August 3, 1990, Linsley tested Brady's construct in a standard assay with B7 expressing cells and observed that CTLA4 bound B7. In fact, CTLA4 bound B7 with higher affinity than CD28, which was critical to its utility as an immune modulator because that allowed it to disrupt the CD28/B7 interaction.

Plaintiffs argue there are discrepancies in the dates which appear in Mr. Brady's and Dr. Linsley's laboratory notebooks. There was no uniform practice at Bristol of writing in the "issued on" date when the notebook was actually given to the scientist. Mr. Brady returned all of his notebooks on January 2 and 3, 1995, in preparing to leave his employment. On all of his notebooks, the "issued on" date is written in the same color ink and handwriting as the "returned on" date. In four of the notebooks, the "issued on" date as written is later than the date of the first experiment in the book. The "issued on" dates were written in when the books were returned as a group, not when they were actually given to Mr. Brady, and the clerk who wrote them in made several errors. The notebook log also contains the same errors, such as the "issued on" date does not correlate with the numerical sequence of the notebooks. The Court finds no foul play as to the notebook dates, and considers the matter of back-dating to be a non-issue.

CONCLUSIONS OF LAW

The inventors named on United States patents are presumed correct. See Hess v. Advanced Cardiovascular Sys., Inc., 106 F.3d 976, 980 (Fed.Cir.1997). To overcome the presumption of correctness, plaintiffs must prove by clear, convincing and corroborated evidence that Dr. Thompson is the sole or joint inventor of the subject matter claimed in the patents-in-suit. See, Ethicon, Inc. v. U.S. Surgical Corp., 135 F.3d 1456, 1461 (Fed.Cir.1998).

The joint inventorship inquiry focuses on identifying those among the group who contributed to the conception of the subject matter of the claims at issue. See, e.g., Kimberly-Clark Corp. v. Procter & Gamble Dist., Co., Inc., 973 F.2d 911, 916 (Fed.Cir.1992). "Conception is ``the touchstone of inventorship,' and each joint inventor must generally contribute to the conception of the invention." Board of Education v. American Bioscience, Inc., 333 F.3d 1330, 1337 (Fed.Cir.2003) (citations omitted). "Conception" is defined as "the formation in the mind of the inventor, of a definite and permanent idea of the complete and operative invention, as it is hereafter to be applied in practice." Id."An idea is sufficiently ``definite and permanent' when ``only ordinary skill would be necessary to reduce the invention to practice, without extensive research or experimentation.'" Id.

To make a case for joint inventorship, the inventor must show that he made "a contribution to the claimed invention that is not insignificant in quality, when that contribution is measured against the dimension of the full invention, and [did] more than merely explain to the real inventors well-known concepts and/or the current state of the art." Pannu v. Iolab Corp., 155 F.3d 1344, 1351 (Fed.Cir.1998). Each inventor must work on the "same subject matter and make a contribution to the inventive thought and to the final result [.]" Monsanto Co. v. Kamp, 269 F.Supp. 818, 824 (D.D.C.1967). The contribution of each inventor, however, need not be the same type or amount, and a joint inventor must make such a contribution only to one claim of a patent. Id. at 824. A joint inventor need not have had the definite and permanent conception of the full invention — otherwise he would be the sole inventor.

To prove inventorship, there must be corroborating evidence to support a claim of inventive contribution. See Sandt Tech., Ltd. v. Resco Metal & Plastics Corp., 264 F.3d 1344, 1350 (Fed.Cir.2001). In determining whether corroboration exists, courts apply a "rule of reason" analysis under which the Court must evaluate "all pertinent evidence ... so that a sound determination of the credibility of the inventor's story may be reached." Id. at 1367; Price v. Symsek, 988 F.2d 1187, 1194 (Fed.Cir.1993). The evidence of corroboration must be considered as a whole, not analyzed piecemeal. See Price, 988 F.2d at 1195-96. However, there must be independent corroboration of all relevant aspects of the asserted contribution. See Acromed Corp. v. Sofamor Danek Group, Inc., 253 F.3d 1371, 1379 (Fed.Cir.2001).

Corroborative evidence does not require the "actual witnessing of the conception or reduction to practice of the invention by one who understands the substance of the interaction in order to be adequate[.]" Berges v. Gottstein, 618 F.2d 771, 774 (Cust. & Pat.App.1980). Nor must the corroboration be proven by documentary evidence. See Loral Fairchild Corp. v. Matsushita Elec., 266 F.3d 1358, 1364 (Fed.Cir.2001) ("[T]he district court's insistence upon ``documentary evidence' to corroborate test results of Dr. Amerlio's claimed reduction to practice was erroneous as a matter of law."). As this case has made clear, scientists, at the time of this collaboration, were not in the general habit of putting the terms of their relationships with one another in writing. The legal community has a hard time relating to this manner of doing business, but has attempted to come up with workable standards by which to analyze dealings among scientists. The Court of Appeals for the Federal Circuit has stated that "it is a rare case indeed that some physical record (e.g., a written document such as notes, letters, invoices, notebooks, or a sketch or drawing or photograph showing the device, a model, or some other contemporaneous record) does not exist." Woodland Trust v. Flowertree Nursery Inc., 148 F.3d 1368, 1373 (Fed.Cir.1998). Circumstantial evidence, including oral testimony of those to whom the putative inventor disclosed his conception, may provide sufficient corroboration. See, e.g., Cooper v. Goldfarb, 154 F.3d 1321, 1330 (Fed.Cir.1998). However, oral testimony must be viewed carefully as it "is peculiarly untrustworthy, particularly if the testimony is rendered after lapse of years ...." See Finnigan Corp. v. Int'l Trade Comm'n, 180 F.3d 1354, 1368 (Fed.Cir.1999), quoting Deering v. Winona Harvester Works, 155 U.S. 286, 15 S.Ct. 118, 39 L.Ed. 153 (1894). An alleged rival Inventor's testimony, standing alone, is not sufficient to support a claim of inventorship — it must be corroborated, preferably by contemporaneous documentary evidence, or the testimony of a disinterested third party witness. See, Price v. Symsek, 988 F.2d 1187, 1194-95 (Fed.Cir.1993).

"The law does not impose an impossible standard of ``independence' on corroborative evidence by requiring that every point of a reduction to practice be corroborated by evidence having a source totally independent of the inventor; indeed, such a standard is the antithesis of the rule of reason." Cooper, 154 F.3d at 1331 (citation omitted). (corroborative evidence included contemporaneous documents and testimony about the noted experiments by the inventor and others). Circumstantial evidence of an independent nature can satisfy the corroboration requirement. In re Jolley, 308 F.3d 1317, 1325 (Fed.Cir.2002).

"Every putative inventor must nonetheless provide corroborating evidence of any asserted contributions to the conception of the invention. Like conception of the entire invention, a contribution to conception is a mental act which cannot be accurately verified without corroboration." Fina Oil and Chem. Co. v. Ewen, 123 F.3d 1466, 1474 (Fed.Cir.1997). "[T]he temptation for even honest witnesses to reconstruct, in a manner favorable to their own position, what their state of mind may have been years earlier, is simply too great to permit a lower standard." Hess, 106 F.3d at 980.

Plaintiffs argue that Thompson is a joint inventor because he provided the practical utility for both CTLA4-lg and the antibody to CTLA4, which provided the basis for patentability of all of these patents. That utility was based on his conception that CTLA4 binds B7 and by blocking that binding, the T-cell and B-cell interaction could be regulated. Plaintiffs argue that Thompson provided the conception and contribution on which Bristol based the patentability of these claims and the basis on which the Examiner allowed the claims. Brenner v. Manson, 383 U.S. 519, 529, 86 S.Ct. 1033, 16 L.Ed.2d 69 (1966). On the other hand, defendant argues that Linsley discovered the practical utility of CTLA4-lg when he determined on August 3, 1990 that it bound with high affinity to B7.

Thompson first brought CTLA4, then an orphan gene without a known function, to the attention of the collaborators in April or May 1989 in the context of their studies of the CD28 pathway. However, informing the named inventors of publicly available information is not conception of an invention or a contribution to an invention, and thus does not make one an inventor. See Hess v. Advanced Cardiovascular Sys., Inc., 106 F.3d 976, 981 (Fed.Cir.1997). The published literature from Brunet, Golstein, Dariavach, and Hunkapillar and Hood,^[4] pointed out that CTLA4 was similar to CD28, was classified in the same subfamily of the lg superfamily, and might bind another superfamily member, and suggested further study in this area.

At this same time, in April or May of 1989, Thompson testified that he informed Linsley that CTLA4 and CD28 were related molecules and of his hypothesis that CTLA4 is structurally and functionally related to CD28. If true, this states no more than was disclosed in the literature. For example Brunet reported that CTLA4 and CD28 were related molecules, sharing structural similarities, and suggesting further study to determine if they performed related functions. None of the collaborators were aware of CTLA4 or its potential functional relationship to CD28 prior to this disclosure. This is irrelevant, however, because the existence of CTLA4 and its relationship to CD28 was published in the literature by the time Thompson brought it to Linsley's attention.

By May 1989, Thompson's laboratory had completed much of the genomic organization of CD28. By letter dated May 15, 1989, Thompson sent a copy of this genomic organization of CD28 to Linsley. This fact is irrelevant to whether Thompson made any contribution to the CTLA4-related inventions claimed in the patents-in-suit. There is no record evidence that the letter was even received by Linsley. More importantly, the letter makes no reference to CTLA4, lg-fusion protein technology, gene duplication, or that B7 was the natural ligand for CD28. No one could have hypothesized that a B7 binding molecule could potentially disrupt the co-stimulatory pathway until Dr. Linsley discovered that B7 bound CD28, which did not happen until December 2, 1989.

On May 16, 1989, after reading the Golstein article, Linsley downloaded the CTLA4 gene sequence on his computer. He then continued his experiments to determine the natural ligand of CD28, which he established to be B7 on December 2, 1989. Next he proceeded to develop soluble forms of molecules to determine their effect on the CD28 pathway, first B7-lg, then CD28-lg, then CTLA4-lg, because of the published similarity between CD28 and CTLA4.

In December, 1989, Thompson compared the pre-publication results from his laboratory of the genomic organization of CD28 and the previously published genomic organization of CTLA4 and concluded that these two molecules resulted from a recent (in evolutionary time) gene duplication. Thompson never recorded this conclusion in any written record. At his deposition, Thompson testified that he told Linsley, possibly in late April, May or June of 1990, that CTLA4 would bind a B7-related molecule. Thompson recanted his deposition testimony and said for the first time on rebuttal at trial that this occurred in 1989. Thompson's testimony is uncorroborated and vacillating, lacks foundation, and is not clear and convincing evidence. Ledbetter's testimony of what Linsley told him "sometime in 1989" was imprecise and unreliable because of his admitted weakness on dates, the fact he was not a participant in the alleged communication, his inability to recall anything else about the conversation, and the fact that Thompson never expressed any belief directly to him.

In November and December 1989, Thompson and Lee were discussing drafts of the manuscript of the article on the genomic organization of CD28, which was then submitted for publication on January 17, 1990. By September 1989, Thompson testified that he knew that Bristol had strong indications that B7 bound CD28. There is no corroboration for this statement, which was made by Thompson for the first time during rebuttal at trial. On December 2, 1989, Linsley confirmed that CD28 binds B7. On December 7, 1989, Thompson's laboratory submitted the Turka Article to which the note added in proof disclosing CD28-B7 binding was to be appended. At his deposition, Thompson testified that discussions about adding the note added in proof began "sometime about a week after the paper was accepted" which would be after December 8, 1989. The fact that the note added in proof was contained in the article published in March 1990 corroborates Thompson's testimony that discussions took place between his laboratory and Bristol. However, such evidence does not corroborate that they discussed CTLA4, which had nothing to do with the added note.

At his deposition, Dr. Thompson testified that the first he could say under oath that he made the suggestion that the Bristol scientists should make a soluble form of CTLA4, including CTLA4-lg, was at the July 23, 1990 meeting at Oncogen. Ledbetter's reaction was that he was hearing it for the first time, not "again". All of the evidence, including Ledbetter's testimony, is that by this time the Bristol scientists already had the idea and were in the process of making a CTLA4-lg fusion protein to determine if it bound B7.

After the meeting, Ledbetter had a discussion with Linsley that he should focus on making CTLA4-lg. Brady had already made CD28-lg and his laboratory notebook number 1875 corroborates that he began work prior to the July 23, 1990 meeting to make a CTLA4-lg fusion protein. On July 18, 1990, Brady was able to locate a source of CTLA4. The fact that Brady did the experimental work to construct CTLA4-lg after the July 23, 1990 meeting is irrelevant. All Thompson expressed was an idea, and there is an abundance of corroborative evidence that Bristol had the idea before Thompson communicated it to them.

'095 Patent

Plaintiff argues that Dr. Thompson is the sole inventor of the '095 Patent. The '095 Patent contains two claims. Claim one reads:

A CTLA4 lg fusion protein which binds the B7 antigen, said protein having a first amino acid sequence consisting of the extracellular domain of CTLA4 and a second amino sequence consisting of the hinge, CH2 and CH3 regions of a human immunoglobulin molecule.

This claim consists of two elements. The first element is the function of CTLA4-lg, that it binds B7. The second element is the particular lg fusion protein consisting of the extracellular domain of CTLA4 and the constant region of the lg fused to it as a "tail".

The Court has held that regardless of when Dr. Thompson first conceived that CTLA4 binds B7, by the time he disclosed this conception to Bristol they already had the same idea and had begun working to make CTLA4-lg to test for binding with B7. The Court has also found that the work required to construct CTLA4-lg was not routine in 1990. Even if the pieces of the compound, CTLA4-lg, were known in the art, and even if the art taught how to put those pieces together, the person who conceives a compound must conceive the precise claimed combination of pieces. American Bioscience, 333 F.3d at 1340. There is no suggestion that Dr. Thompson concelved of the complete molecule claimed with all of its component constituents. Accordingly, Dr. Thompson is neither the sole inventor of the '095 Patent nor a joint inventor.

'131 Patent

The '131 Patent claims binding of the extracellular domain of CTLA4 with B7 as a means of regulating T-cell interactions. The core of the claims of the '131 Patent is Dr. Thompson's conception that CTLA4 binds B7 and acts as an immunosuppressant. As held above, the record is devoid of any clear and convincing evidence that Dr. Thompson communicated the conception to Bristol that the function of CTLA4 is to bind to B7 and act as an immunosuppressant, prior to the time that Bristol already had that idea. Dr. Thompson is neither the sole inventor of the '131 Patent nor a joint inventor.

Remaining Patents

Plaintiffs argue that Dr. Thompson should be named as a joint inventor on the remaining patents-in-suit. However, plaintiffs have not established that Dr. Thompson made any inventive contribution to conceiving, designing or making any of the molecules, proteins, plasmids or host vector systems recited in the '795 Patent. Nor have plaintiffs established that Dr. Thompson made an inventive contribution to conceiving, designing, or making any of the subject matter recited in the claims of the '796 Patent. The '579 Patent claims a method for regulating T-cell interactions or treating patients by using a soluble CTLA4 by contacting B7-positive cells with a CTLA4 fusion protein. Plaintiffs have not established by clear, convincing and corroborated evidence that Dr. Thompson made a significant contribution to the claimed invention, for the same reason that they have failed to establish that Dr. Thompson should be a named inventor on the '131 Patent. Finally, the claims of the '510 and '318 Patents require the use of a monoclonal antibody that reacts with the CTLA4 receptor to inhibit the binding of B7 to that receptor, Plaintiffs have not established by clear, convincing and corroborated evidence that Dr. Thompson made an inventive contribution to the claims of the '318 or the '510 Patents because plaintiffs have provided no evidence that Dr. Thompson conceived of the use of the claimed antibodies for the recited purpose, or contributed in any way to the design or making of such monoclonal antibodies.

Plaintiff characterizes Thompson as a joint inventor because he selected one gene, CTLA4, out of the thousand possible genes and conceived of its function in the invention. See Boots Co. v. Analgesic Assocs., 1993 WL 37123, 26 USPQ2d 1144, 1146 (S.D.N.Y.1993). However, it is a soluble form of the protein, not the gene, that is referred to in the claims. Golstein discovered CTLA4, not Thompson. Linsley discovered that B7 was a natural ligand for CD28 and CTLA4, not Thompson. Linsley designed the primers used to make the claimed CTLA4-lg fusion protein, not Thompson. Brady made the claimed CTLA4-lg fusion protein, not Thompson. Linsley determined that CTLA4-lg bound B7 with a higher affinity than it bound CD28, not Thompson. If Thompson suggested anything, it was either already in the literature or already conceived by the Bristol scientists.

CONCLUSION

Plaintiffs have failed to meet their burden to prove by clear, convincing and corroborative evidence that Dr. Thompson is a sole or joint inventor of any of the patents-in-suit. Judgment shall enter on behalf of defendant.

NOTES

[1] Lee et al., "The Genomic Organization of the CD28 Gene: Implications for the Regulation of CD28 mRNA Expression and Heterogeneity," Journal of Immunology, Vol. 145, No. 1, pp. 344-352 (1990).

[2] Turka et al., "CD28 is an Inducible T Cell Surface Antigen that Transduces a Proliferative Signal in CD3+ Mature Thymocytes," Journal of Immunology, Vol. 144, No. 5, pp. 1646-1653 (1990).

[3] LaFage-Pochitaloff, et al., "Human CD28 and CTLA4 lg Superfamily Genes are Located on Chromosome 2 at Bands q33 — q34" Immunogenetics Vol. 31, No. 3 pp. 198-201 (1990).

[4] Brunet, et al., "A Different Molecular Biology Search for Genes Preferentially Expressed in Functional T Lymphocytes: The CTLA Genes," Immunological Review No. 103, pp. 21-36 (1988).

Dariavach, et al., "Human lg Superfamily CTLA-4 Gene: Chromosomal Localization and Identity of Protein Sequence Between Murine and Human CTLA-4 Cytoplasmic Domains," Eur. J. of Immunology Vol. 18, pp.1901-1905 (1988).

Hunkapillar and Hood, "Diversity of the Immunoglobulin Gene Superfamily," Advances in Immunology Vol. 44, pp. 1-63 (1989).