United States: Investing In University Spin-Out Companies

This issue of the Tech Transfer Update was prepared for "First Look L.A.," a showcase of technology from California Institute of Technology, the University of Southern California and the University of California Los Angeles, held November 7, 2007 in Los Angeles, before an invited audience of venture capitalists and angel investors.

1. Unique Legal Issues Affecting University Spin-outs
2. Bayh-Dole Act Background
3. March-in Rights
4. Preference for United States Manufacturing
5. Inventor Royalty Share
6. Trade Secrets and University Policies
7. Conclusion

1. Unique Legal Issues Affecting University Spin-outs

There are some special legal considerations associated with investing in a university spin-out company. Some of these arise as a result of the Bayh-Dole Act, a federal law that governs intellectual property rights in inventions stemming from federally funded research. Investors and investors' legal counsel generally conclude that these legal issues are not an obstacle to investment, and that the opportunity for reward in university spin-outs outweighs the risks. But it is useful for investors and their legal counsel to approach investment in a university spin-out with a basic understanding of relevant provisions of the Bayh-Dole Act and also typical university policies. This can save a lot of unnecessary headache during the transaction negotiation and documentation process and afterwards. This memorandum is intended to convey that understanding, and will discuss "march-in" rights, preference for U.S. manufacturing and the inventor's royalty share, all of which stem from the Bayh-Dole Act, and also the interaction between trade secret law and university policies on free flow of information.

2. Bayh-Dole Act Background

The Bayh-Dole Act is a federal law enacted in 1980. Before its passage, there was a widespread perception that too little federally funded research was being commercialized. Patent policies varied among different federal agencies, and commercial enterprises generally did not receive exclusive rights to government-owned patents. A prospective corporate licensee of a federally funded invention was faced with the unappealing prospect of high-risk, long-term investment in the transition from laboratory bench to market, after which other companies that did not undertake the same process would be able to step up and receive competing licenses. The principal purpose of the Bayh-Dole Act was to promote the utilization of inventions arising from federally supported research, by giving incentive to private enterprises to commercialize those inventions.¹ This includes the possibility of exclusive licenses.

Under the Bayh-Dole Act, a university can elect to patent and own an invention growing out of government funded research.² The university can grant licenses, either non-exclusive or exclusive, covering the invention. The company receiving the license, and its investors, then have an incentive to invest in development of products based on the invention, with an expectation of a secure intellectual property position. The company receiving such a license may be either an established company or a startup venture formed to exploit the technology.

3. March-in Rights

Under the Bayh-Dole Act, a federal agency that funded an invention may require the patent holder to grant a license to a responsible applicant on reasonable terms. If the patent holder refuses, then the federal agency can grant a license itself.³ This retained federal power, referred to as "march-in rights," is initially worrisome to investors' counsel if they have not dealt with university spin-outs before, but it need not be. The concern is that the spin-out company that exclusively licenses a federally funded invention from a university will not get the real patent monopoly that an exclusive license ordinarily implies. But solidly established precedent makes it extremely unlikely that march-in rights would ever be exercised in a manner that would trouble an investor. In fact, march-in rights have never been exercised by the government in the 26 years since the Bayh-Dole Act was passed.

Criteria for Marching In. The law directs federal agencies not to exercise march-in rights unless the action is necessary:

• Because the patent holder is not taking effective steps to achieve practical application of the invention,

• To alleviate health or safety needs which are not reasonably satisfied by the patent holder,

• To meet requirements for public use specified by federal regulations not reasonably satisfied by the patent holder; or

• Because an exclusive licensee has failed to give preference to U.S. manufacturing where that would be required.⁴

The National Institutes of Health ("NIH") have been called on to apply these criteria. Written decisions of the NIH, discussed below, display a keen awareness that exercise of march-in rights could disrupt the incentive for commercialization of federally funded research, and thereby undermine attainment of the principal purpose of the Bayh-Dole Act.

CellPro 1997 Petition to NIH. The first published decision regarding march-in rights was rendered by NIH in the CellPro case. In 1984, Dr. Curt Civin, in the course of research funded by NIH at Johns Hopkins University, devised a technology for isolation of stem cells. The technique involved taking an antibody that targets stem cells that can generate blood cells, and attaching it to a magnetic substance. The technology had both therapeutic and diagnostic applications. Baxter Healthcare obtained an exclusive license to the patent. Another company, CellPro, developed a competing stem cell selection device and obtained FDA approval for use of its product. Baxter sued CellPro for patent infringement and won. Baxter refused to grant CellPro a license. CellPro asked NIH to invoke march-in rights. CellPro's argument was that it had the only FDA-licensed product, since Baxter's product had not yet been approved. CellPro's petition stated that if it were not allowed to sell its product, patients would suffer. CellPro asked NIH to require Baxter to grant a license to CellPro.

The NIH refused to grant CellPro's petition. The thoughtfully written decision gives several reasons for this. First, Baxter was not sitting on the technology. They were working on getting FDA approval for their own product. Second, the Baxter product was in fact widely available to patients through clinical trials. Baxter had also pledged that where the Baxter product was not yet available, Baxter would allow CellPro to sell the CellPro product until the Baxter product was available. Third, Baxter had offered CellPro a license prior to the lawsuit. CellPro had declined to pay for the license, instead taking the position that it did not infringe and did not require a license and rolling the dice in litigation. NIH was reluctant to intervene in what it perceived as a commercial dispute.

Most importantly, NIH understood that to intervene would threaten future investments in new university inventions. The decision stated:

"We are wary, however, of forced attempts to influence the marketplace for the benefit of a single company, particularly when such actions may have far-reaching repercussions on many companies' and investors' future willingness to invest in federally funded medical technologies. The patent system, with its resultant predictability for investment and commercial development, is the means chosen by Congress for ensuring the development and dissemination of new and useful technologies. It has proven to be an effective means for the development of health care technologies. In exercising its authorities under the Bayh-Dole Act, NIH is mindful of the broader public health implications of a march-in proceeding, including the potential loss of new health care products yet to be developed from federally funded research."⁵

NIH Director Letter to Nader. Consistent NIH decisions subsequent to the CellPro case have now firmly entrenched the principle that the federal government is very reluctant to exercise march-in rights. In 1999, NIH Director Harold Varmus wrote to Ralph Nader, rejecting a request for NIH to license the World Health Organization to use U.S. government funded medical inventions. In the letter, he stated:

"As a practical matter, it is reasonable to assume that companies will not undertake the development costs of these inventions if they believe the Government will readily allow third parties to practice the inventions."⁶

Abbott Labs / Norvir. In 2004, the NIH rendered another march-in decision, this one pertaining to an HIV / AIDS drug called Norvir, for which a patent was held by Abbott Laboratories. Abbott had developed the drug partially with federal funds.

As noted above, one basis for exercise of march-in rights under the Bayh-Dole Act would be failure of the patent holder to take effective steps to achieve practical application of the invention. Practical application is defined in the law to mean that the invention is utilized and its benefits available to the public on reasonable terms.⁷

The drug Norvir was approved by the FDA and widely prescribed. The issue in that case was the price. The drug was patented, and expensive, so the issue in the case was whether the benefits were available to the public "on reasonable terms." Members of Congress and the public petitioned NIH to exercise march-in rights. The NIH ruled that "because the market dynamics for all products developed pursuant to licensing rights under the Bayh-Dole Act could be altered if prices on such products were directed in any way by NIH . . . the extraordinary remedy of march-in is not an appropriate means of controlling prices."⁸

Pfizer / Xalatan. The Xalatan case involved a glaucoma treatment developed jointly by Columbia University and Pharmacia, which was later bought by Pfizer. This product was FDA approved, and covered by a patent exclusively licensed by Columbia University to Pfizer. The price of the drug was higher in the United States than in Canada or Europe. Members of the public expressed concern over that pricing and petitioned the NIH to exercise march-in rights. In a written decision very similar to the Norvir case, the NIH declined to intervene.⁹

Conclusion. The CellPro, Norvir and Xalatan cases, together with the Nader letter and public NIH pronouncements, firmly establish a precedent of extreme caution, if not outright hostility, regarding the exercise of march-in rights. The U.S. government has never exercised march-in rights. One can imagine exercise of march-in rights consistent with established precedent only in extreme circumstances. Perhaps, if a company licensed a useful medical product, but then deliberately refused to make it available, thereby depriving the public of its benefits, exercise of march-in rights could be justified under existing precedent. Similarly, one can imagine that if a company developed a useful defense technology with federal funds, and then refused to (or was unable to) make it available to the United States military, that would be a scenario in which the Defense Department might exercise march-in rights. But companies that license a technology normally do so with the intention of making it available for sale, so these outlying scenarios do not pose a practical threat. Investors contemplating financing a university spin-out have generally concluded that they can do so without concern that the federal government would exercise march-in rights in a way that would interfere with their expectations.

4. Preference for United States Manufacturing

A second feature of the Bayh-Dole Act that investors need to understand is the preference for U.S. manufacturing. If a university grants an exclusive license to a federally funded invention, the Bayh-Dole Act requires the licensee to manufacture substantially in the United States. An exception applies if domestic manufacture is not commercially feasible, or if the university tried unsuccessfully to find a licensee that was likely to manufacture in the U.S. To take advantage of one of these exceptions, a waiver must be obtained from the agency that funded the research.¹⁰

Unlike in 1980 when the Bayh-Dole Act was passed, in 2007 most manufacturing of electronic and consumer products is done outside the U.S., where costs are dramatically lower. This is less true in the life science field. For industries where U.S. manufacturing is not practical, the most likely solution is to seek a waiver of the requirement. If one assembles convincing evidence that U.S. manufacturing is not commercially feasible, a waiver can ordinarily be obtained. There is some difference between the federal agencies, with NIH having a reputation for a streamlined waiver process, and a willingness to grant waivers fairly freely, particularly if there is reason to believe that foreign manufacturing will allow a medical product to be made available to patients at a lower cost. The Department of Defense, not surprisingly, may be somewhat less quick to agree to foreign manufacture, particularly for a sensitive military product. If DOD is the funding agency, an investor might reduce risk by requiring the waiver to be obtained prior to funding the investment, though this could involve substantial delay.

An application for waiver of the requirement of substantial manufacturing in the U.S. should address the following points:

• Does anyone in the industry manufacture in the U.S.? Where are competing products produced?

• What are the relative costs of U.S. versus foreign manufacture of the product in question?

• Will the licensee do some manufacturing in the U.S. even if not enough to be considered "substantial," such as final assembly?

• Is there other economic value to the U.S. economy, despite the absence of U.S. manufacturing, such as additional jobs in the U.S. in research and development, design, or sales and marketing; taxes to be paid by the licensee for sales in the U.S., etc.?

• If applicable, are there restrictions in the license so that foreign-manufactured products are only authorized for foreign sales, and the U.S. market is still reserved for U.S. manufacturing?

• Is there any ancillary consideration that makes the granting of a license beneficial to the U.S. public? (For example, is it an environmentally responsible technology?)¹¹

Addressing these matters thoroughly and with detailed documentation will make it easier for the agency to approve a waiver application, and likely speed the process. Less thorough applications are likely to result in follow-up questions and delay. It can be helpful to contact the agency personnel in advance to discuss the proposed submission.

A couple of other avenues are available if one is not willing to go through the waiver process. The U.S. manufacturing requirement only applies to an exclusive license. Most of the time a spinout company will not be interested in a non-exclusive license. But in some cases a company may be satisfied with a license that allows other licensees, but only in a very limited number (one or two). One could take the position that this is not an exclusive license, though the issue is not free from doubt.

Another approach involves interpretation of the requirement to manufacture "substantially in the United States." There is no clear percentage requirement, nor is there guidance from the Department of Commerce or other sources on what "substantially" means. For electronic products, one might have parts such as circuit boards made abroad, but then brought to the U.S. for final assembly, software installation and quality control. If sufficiently key elements take place in the U.S., arguably the requirement for "substantial" manufacture in the U.S. is met.

The two approaches described above leave some risk of challenge regarding whether the license was truly exclusive or whether the product was really substantially manufactured in the U.S. Failure to meet the U.S. manufacture requirement where applicable would allow the agency that funded the research to exercise march-in rights (discussed in Section 3 above). Unless the federal agency is likely to exercise march-in rights, the practical risk may not be great. A federal district court in New Jersey has determined that violation of the U.S. manufacturing requirement by a licensee does not result in any adverse consequences to the exclusive license unless and until the applicable agency invokes its march-in rights.

This decision about the U.S. manufacturing requirement arose out of Ciba-Geigy Corp. v. Alza Corp., 804 F.Supp. 614 (1992). In that case, Ciba-Geigy had an exclusive license for a nicotine patch from the Regents of the University of California. Ciba-Geigy sued Alza, claiming that Alza's product Nicoderm infringed the Regents' patent. Alza counterclaimed that Ciba-Geigy's exclusive license from the university was not valid because Ciba-Geigy had been manufacturing its product in Germany, in violation of the U.S. manufacturing requirement. The court determined that Alza could not defend against an infringement claim based on the failure of Ciba-Geigy to meet the U.S. manufacturing requirement. The court ruled that failing to manufacture in the U.S. does not automatically invalidate an exclusive license nor convert it to a non-exclusive license, so long as the government agency that funded the invention does not invoke its march-in rights. The funding agency had discretion as to whether or not to exercise march-in rights, and unless and until it chose to do so (which it had shown no interest in doing), the license was unaffected.

In conclusion, there are generally satisfactory ways of dealing with the Bayh-Dole U.S. manufacturing requirement. The safest is to submit a thorough waiver application, documenting why U.S. manufacture is not commercially feasible, and also pointing out other benefits to the U.S. economy of granting an exclusive license to the spin-out company. If there is a reason to believe that a waiver might be harder to obtain in a particular case, such as in the case of a sensitive military technology, consideration might be given to requiring obtaining a waiver as a condition of funding an investment.

5. Inventor Royalty Share

A third feature of the Bayh-Dole Act that is important for a university spin-out investor to understand is the inventor's royalty share. When a university receives royalty payments for a licensed federally funded invention, the university is required by law to share a portion of those payments with the inventor or inventors. This applies not only in the case of an established corporate licensee, but also where the licensee is a spin-out company of which the inventor is a founder and perhaps board member and / or officer.¹²

The amount of the inventor's share is not set by law, and varies from one institution to the next. Platzer v. Sloan-Kettering Institute for Cancer Researchwas a challenge to the Sloan-Kettering royalty policy based on the Bayh-Dole Act.¹³ Dr. Platzer and four others had invented a method of purifying granulocyte colony stimulating factor. G-CSF stimulates the production of white blood cells, and thereby enables cancer patients to tolerate higher dosages of chemotherapeutic drugs. Sloan-Kettering Institute received a $50 million balloon payment for licensing this technology. The Sloan-Kettering royalty policy gave the investor 25% of annual gross proceeds from $0 - $50,000; 15% of annual gross proceeds from $50 - $150,000; 10% of annual gross proceeds from $150,000 - $300,000, and 5% of annual gross proceeds over $300,000. Because this case involved a large royalty payment that was all made in a single year, the declining percentage structure meant that the smallest 5% rate applied to almost all of the payment, so that the total inventor's share was only $2,527,500 of the $50 million. There were five inventors, so this in turn had to be split among them, leaving $505,500 for Dr. Platzer, which he evidently regarded as insufficient for his role in an invention that had brought $50 million to the Institute.¹⁴

The Court rejected Dr. Platzer's challenge, concluding that while the Bayh-Dole Act requires the institution to share royalties with the inventor, it does not specify any minimum percentage. According to the Platzer decision, in adopting the Bayh-Dole Act, Congress was concerned with promoting commercialization of inventions and reinvesting funds into research, not furthering the private interests of inventors. The Court stated that Congress intended that organizations share royalties in accordance with their own usual policies, as established by market forces.

Investors in university spin-outs should review the patent policy of the relevant university in order to understand the incentives of the inventor, if the inventor will have an ongoing role in the spin-out. These policies are generally set out in full at the university's website. If the inventor is an officer of the spin-out company, there is a potential conflict between the interests of the inventor and the investors that needs to be managed. Depending on the equity stake of the inventor in the company, it is possible that he or she would benefit more from the company paying a higher royalty to the university rather than a lower royalty. Sometimes the investor will request that the inventor assign his inventor's share back to the company, but this may not be necessary, and will not necessarily eliminate a skewing of incentives. For example, it could create a situation in which the inventor is better served if an invention is licensed to another company instead of the spin-out company. The best solution may be for a company representative other than the inventor to lead negotiation of the license terms with the university. That can also address potential conflicts between the interests of the inventor and investors with respect to other license terms, including duration, breadth of field, future inventions and restrictions on assignment.

6. Trade Secrets and University Policies

Some form of intellectual property protection is a very important factor in the investment decision of most investors in early stage enterprises. Someone investing in validating a new technology will want assurance that it will not be easy for others who have not made the same investment to step in and compete once the viability of the enterprise has been demonstrated. For most startup companies, the main IP protection is either through patents or through trade secrets, or possibly through patents for some assets and trade secret protection for others.¹⁵ Each of patent protection and trade secret protection has advantages and disadvantages, and companies will ordinarily choose the most suitable form of IP for a particular technology. Only patent protection, not trade secret protection, will usually be available for inventions growing out of university research. This is because all universities have a strong policy favoring an open flow of information that is antithetical to trade secret protection. The initial IP protection strategy for a university spin-out almost always takes the form of an exclusive license from the university on one or more patents or patent applications. Most investors, and their legal counsel, conclude that patents will adequately protect their investment without the availability of trade secret protection. But for those new to university spin-outs, this requires some education. The IP license will have some different features from what one would see if the licensor was a commercial enterprise. If there is a consulting arrangement between the spin-out and a university faculty member, the consulting agreement will also probably need to be different from those in the form files of the investor's counsel (unless that counsel has previous experience with university spin-outs). Of course, the spin-out company can still obtain trade secret status for inventions that grow out of its own internal research.

The definition of a trade secret can vary from state to state¹⁶, but a key element is that the owner has to keep the information secret. Many states have adopted the Uniform Trade Secrets Act, which provides protection for information that has actual or potential independent economic value stemming from its secrecy, and that has been the object of reasonable efforts designed to maintain its secrecy. Under common law, courts have adopted the Restatement of Torts' definition of a trade secret: "[a] trade secret may consist of any formula, pattern, device or compilation of information which is used in one's business, and which gives . . . an opportunity to obtain an advantage over competitors who do not know or use it." Information that is general knowledge in an industry or is readily available by examining a product cannot be a trade secret. Trade-secret subject matter can include chemical formulae, industrial processes, know-how (methods and techniques), products, customer lists and information, sources of supply, and business information.

In order to qualify information as a trade secret, the owner must take reasonable steps to keep the information secret. Such steps may include, for example, confidentiality agreements, restrictions on access, physical locks on file cabinets that contain secret information, computer passwords, policies restricting transfer of information outside of a company and the like. Patent and trade secret protection may not exist for the same technology, because publication is a requirement of obtaining a patent.

As noted above, there are advantages and disadvantages to patent and trade secret protection. Trade secret protection may cost less than patent protection, and is available immediately whereas patent protection involves the delay of prosecution in the patent and trademark office. Trade secrets in principle can last indefinitely (if the secret can be maintained), whereas patents last only 20 years. Patents require publication, which may invite infringement and require formal legal enforcement. With trade secrets, by contrast, if they are successfully maintained, the competitor does not know how to infringe even if they wanted to do so. A key disadvantage of trade secret protection is the difficulty of maintaining the secret and the possibility and permissibility of reverse engineering. Reverse engineering is accepted by the courts and statutorily defined in many jurisdictions as a proper means by which the non-owner may discover a trade secret. So if a product can be taken apart to determine how to make it, trade-secret protection ends because this is a proper means of discovering the secret. Once properly discovered, the information is in the public domain and no longer protected. Also, trade secret protection, unlike a patent, will not protect against a competitor who, after the original invention, legitimately independently discovers the same secret.

The law's requirement of secrecy in the research process as a prerequisite to trade secret status conflicts with several features of university culture, including:

• A policy in favor of free dissemination of knowledge and open communication

• The drive of individual researchers to publish their findings in order to further career goals (tenure, stature in the scientific community, Nobel and other prizes, etc.)

• Relative autonomy of the functioning of individual faculty members, as compared to industrial researchers

• Non-profit orientation, and value placed on providing public benefit

• In the case of public universities, freedom of information laws that could allow public access, unless a sufficiently strong exemption from that law exists for trade secrets

In light of these policies, universities typically will not agree to license terms that would prohibit publication of research findings. A typical university license or joint development agreement provides a limited delay of publication (30 - 60 days) during which the licensee can determine whether to file for patent protection. Even if universities were willing to adopt restrictions on information flow (which they are not), university laboratories are generally not set up with the kind of procedures that commercial enterprises routinely use to establish trade secret status, such as: preparing and following a written trade secret protection policy; limiting access to the portion of the facility where trade secrets are stored; requiring visitors to sign in and out; requiring all participants to sign confidentiality agreements, etc.

Patent protection is generally the intellectual property strategy of choice for university spin-out companies. For university based research, the alternative of trade secret protection will generally not be available. In most situations, patent protection is considered a sufficient, if not preferable, barrier to competition. The spin-out company can also use trade secret protection for its own internally developed inventions.¹⁷

University technology offers investors a chance to participate from the start in inventions that will change the world. Hence university spin-outs are of great potential interest to investors. Venture capitalists and angel investors contemplating such investments should understand certain special, recurring legal issues applicable to university spin-outs that are different from those encountered with other startup enterprises. These include the Bayh-Dole Act provisions regarding "march-in" rights, preference for U.S. manufacturing and the inventor's royalty share, and also the interaction between trade secret law and university policies on dissemination of information. Most investors, and their counsel, conclude that these issues are not a serious impediment to a profitable investment. March-in rights have never been exercised and appear unlikely to be exercised in any realistic scenario. In those industries in which the Bayh-Dole U.S. manufacturing requirement would pose a problem, there are usually ways around it, most often through obtaining a waiver from the funding agency. The inventor's royalty share may create a conflict of interest, but this can be managed by including a company representative other than the inventor in license negotiations. University policies on free flow of information will require exclusive reliance on patent protection rather than trade secret protection, but patent protection is usually regarded as sufficient to create a very strong intellectual property position for the university spin-out company. In light of the benefit of obtaining exclusive intellectual property rights to innovative research funded by the government rather than the company, on balance, university spin-outs can offer compelling opportunities to investors.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.