The recent decision by the Federal Circuit in Stanford University v. Roche[1] emphasizes how important it is for employers and their counsel to use precise language when drafting clauses governing an employee’s obligation to assign inventions to his or her employer. Stanford University was unable to establish ownership of patents covering inventions developed by an employee even though the employee had signed an agreement containing terms which many would consider more than adequate to protect Stanford’s interests in the inventions. The Stanford case demonstrates how the wrong choice of contractual language can have important and costly consequences, including the loss of valuable patent rights. In this article, we offer some suggestions to help companies and their counsel avoid similar outcomes by carefully managing their employees’ activities with third parties and by carefully drafting assignment provisions in employment agreements.

The patents at issue were developed in part by a Stanford employee while visiting Cetus Corporation to learn a new research method. That employee signed a “Visitor’s Confidentiality Agreement” in which he agreed to assign to Cetus his rights to any inventions made as a consequence of his work there. The Cetus agreement, however, directly conflicted with the employee’s prior Stanford employment agreement, which appeared to require that he assign to Stanford any inventions conceived or reduced to practice during his employment.

The Federal Circuit’s decision in the Stanford case raises a number of issues relevant to agreements allocating patent rights generally. These issues include: (1) the importance of careful drafting of employment and other agreements governing the obligations of researchers and other employees to assign to their employers inventions arising during the course of an employment relationship; (2) the need to consider how to handle situations in which an employee might have conflicting obligations to his or her employer and another organization, whether in the context of a collaboration, a consultancy, or some other type of relationship; and (3) the importance of educating employees regarding what to look for when reviewing the wide variety of agreements executed during the routine course of business.

What Happened in the Stanford Case

The three patents at issue in the case all claimed priority to a common parent application involving a method of monitoring the efficacy of HIV treatments.[2] Stanford researchers performed much of the initial work on this method while working for Stanford until one of them, Mark Holodniy, went to Cetus as a visiting scientist. Three other Stanford researchers, including Holodniy’s laboratory supervisor, Thomas Merigan, were named as co-inventors with Holodniy on some or all of the three patents.

Holodniy joined Merigan’s laboratory at Stanford in 1988 as a Research Fellow in the Department of Infectious Disease. At that time, he signed a “Copyright and Patent Agreement” (the “Stanford CPA”) with Stanford that obligated him to assign any future inventions arising out of his employment to Stanford. According to the terms of the Stanford CPA, Holodniy acknowledged that Stanford entered into “Contracts or Grants” with third parties and that he might “conceive or first actually reduce to practice” various inventions during his employment. His Stanford CPA then also expressly provided that “I agree to assign or confirm in writing to Stanford and/or Sponsors that right, title, and interest in . . . such inventions as required by Contracts or Grants.” As part of the same agreement, Holodniy also expressly promised to “not enter into any agreement creating . . . patent obligations in conflict with this agreement. . . .”[3]

In early 1989, as part of his research project in Merigan’s laboratory at Stanford, Holodniy began to visit Cetus Corporation at Merigan’s direction in order to learn how to perform the polymerase chain reaction (“PCR”). The ultimate goal was to develop a PCR-based assay to monitor the efficacy of HIV treatment. When he first went to Cetus, Holodniy signed a “Visitor’s Confidentiality Agreement” (the “Cetus VCA”). According to its terms, Holodniy agreed that “I will assign and do hereby assign to CETUS my right, title, and interest in each of the ideas, inventions, and improvements” made “as a consequence of” his work at Cetus.[4]

During that same time period, Cetus researchers collaborated with other Stanford inventors besides Holodniy on other aspects of the research project. As part of that collaboration, Merigan, Stanford, and Cetus signed several “Materials Transfer Agreements” allowing Stanford to use certain PCR-related materials and information supplied by Cetus. Those agreements also granted Cetus licenses to technologies created by Stanford using materials and information supplied by Cetus.

Holodniy’s research at Cetus eventually produced a PCR-based assay to measure HIV RNA levels in plasma samples. Subsequent clinical studies at Stanford by the Stanford inventors, including Merigan and Holodniy, confirmed that HIV RNA levels measured by PCR served as a suitable marker for antiretroviral drug efficacy. Those results formed the basis of the patents at issue in the lawsuit.

In December 1991, Roche purchased Cetus’s PCR business, including its agreements with Stanford, and shortly thereafter began to manufacture HIV detection kits using RNA-based assays. In May 1992, Stanford filed the patent application to which the three patents at issue claim priority. In October of 2005, Stanford sued Roche in the Northern District of California, alleging its HIV detection kits infringed the patents at issue.[5]

The outcome of the case turned on the proper interpretation of both the “Visitor’s Confidentiality Agreement” executed during the collaboration between Holodniy and Cetus on the one hand, and Holodniy’s CPA with Stanford on the other hand.

The District Court’s Decision

Roche challenged Stanford’s ownership of the patents, claiming that Roche’s acquisition of Cetus’s PCR assets gave it the right to use the patents under the terms of the Cetus VCA which Holodniy had signed. Roche asserted its theory as a declaratory judgment counterclaim, an affirmative defense, and a challenge to Stanford’s standing to sue for infringement. The district court, however, treated Roche’s defense only as a counterclaim and held that the claim was barred under the applicable California statute of limitations. The court also went on to find all three patents invalid for obviousness. Both parties then appealed to the Federal Circuit.

The Federal Circuit’s Decision

The Federal Circuit began by recognizing that Roche had pled ownership not only in its counterclaim, but also as an affirmative defense. The court held that, since an affirmative defense may be raised at any time—even if the matter alleged would be barred by a statute of limitations if asserted as the basis for affirmative relief—the California statute of limitations did not preclude Roche’s ownership defense. The court then turned to the contractual issue of exactly who owned the patents.[6]

The Federal Circuit’s contractual analysis centered on key differences between the assignment provisions in the Stanford CPA and the Cetus VCA. The Stanford CPA contained language (“agree to assign”) which the court treated as a promise by Holodniy to assign invention rights to Stanford at some time in the future. In direct contrast, the Cetus VCA contained language (“agree to assign and do hereby assign”) which the Federal Circuit treated as an immediate assignment of expectant interests.[7]

The Federal Circuit concluded that (1) Cetus held equitable title from Holodniy at the moment of invention, and (2) Cetus’s legal title to the invention formally vested at the moment the patent application was filed. Because of the difference in the contractual language, the Federal Circuit concluded that Stanford did not own Holodniy’s interest in the invention. According to the court, the terms of the Stanford CPA required that Stanford obtain an actual assignment to perfect its title, even though Holodniy signed the Cetus VCA after he signed the Stanford CPA. In other words, Stanford did not immediately gain title to Holodniy’s inventions on execution of the Stanford CPA, or at the time the inventions were made. The Federal Circuit held that Holodniy’s later assignment to Stanford had no effect because he no longer retained any right, title, or interest in the inventions at that point, having instead conveyed them already to Cetus by the terms of the Cetus VCA.[8]

The Federal Circuit recognized Roche’s ownership interest in the patents derived from Holodniy’s original assignment to Cetus via the Cetus VCA.[9] As a result, the Federal Circuit reversed the district court, finding that Stanford lacked standing to assert its claims of patent infringement because the University could not establish ownership of all right, title, and interest in the patents at issue.

The Lessons of the Court’s Holding

The major lesson of the Stanford case is that language conveying a present assignment of future interests—as with the “agree to assign and do hereby assign” language used in the Cetus VCA—in employment agreements is vastly preferable to language merely promising to convey rights in the future, such as in the Stanford CPA. Agreements that include a present assignment of future interests (like that in the Cetus VCA) enable an assignee to claim equitable title at the moment of invention without the need to have the inventor execute additional assignments later. In contrast, a contract using the “agree to assign” language of the Stanford CPA requires that the inventor execute an additional assignment in order for the assignee to perfect its title in any inventions. There are obvious benefits to be gained by eliminating the need to execute a second document at a later time. And as illustrated by the outcome in the Stanford case, contracts containing a mere promise to assign rights in the future create the risk that an assignee may lose its rights before that later assignment is executed.

An important corollary also is that, whenever an entity has in the past executed agreements containing terms similar to the prospective “agree to assign” language of the Stanford CPA, the employee’s progress on research projects or collaborations should be carefully monitored. All concerned should be prepared to promptly execute the documents required to perfect the assignee’s title. Otherwise, the patent rights may later be lost if an intervening assignment has occurred.

It is worth noting that the Stanford CPA signed by Holodniy also included a clause providing that he expressly promised not to “enter into any agreement creating . . . patent obligations in conflict with this agreement.” However, this clause did not prevent the Federal Circuit from validating the assignment under the Cetus VCA in apparent conflict with Holodniy’s obligations to Stanford. In the court’s decision, it emphasized that the university had either actual or constructive notice of Holodniy’s assignment of rights in the Cetus VCA to Cetus, since Holodniy’s work at Cetus related directly to his work at the university and was known by his Stanford superiors to be taking place.[10] Stanford might have benefited from a stronger and more direct notice requirement in its standard Copyright and Patent Agreement requiring any employee, such as Holodniy, to notify Stanford prior to executing any agreement to assign employee inventions to a third party. Also, Stanford’s position in the litigation would have been strengthened had Stanford inserted a clause in the Cetus/Stanford Materials Transfer Agreements requiring Cetus to provide notice whenever Cetus required or received an assignment of patent rights from any visiting Stanford employee.

The Stanford case also emphasizes the importance of educating employees about potential pitfalls associated with contract clauses that allocate or assign patent rights, and what to look for when agreements contain such clauses. In particular, employees should be instructed never to sign any agreements that contain clauses assigning patent rights, without first obtaining the advice and approval of their employer or their employer’s counsel.

It may also be worthwhile to consider providing a brief seminar on or a written summary of relevant issues accompanied by examples of agreements to all new employees involved in potentially patentable work at the time of hiring. These same employees also should be alerted to the possibility that apparently routine agreements, such as confidentiality or nondisclosure agreements, may unexpectedly include assignment clauses. This information may assist an employee unaccustomed to reviewing legal documents to avoid unwittingly assigning his or her employer’s rights to future inventions to an entity other than the employer.