Panel VI: Intellectual Property and the Public Domain: Sectoral Perspectives

INTRODUCTION

Jorge Goldstein

Stern, Kessler, Goldstein, & Fox

Mr. Goldstein opened the panel by commenting that patents reflect a “grand compromise” in giving inventors incentives to create new technologies—which it is hoped will widely benefit society—by allowing them to privatize the economic returns of an innovation’s benefits. There are questions about what we should privatize, for what length of time privatization should be ensured, and which institutions should benefit from privatization and which should not. The presentations in the panel would, Mr. Goldstein said, reflect the competing forces that run through the patent issue. Dr. Wes Cohen’s presentation would consider how different sectors view patents. Dr. Maryann Feldman’s presentation would focus on patents and universities, and the government-university relationship with respect to patents. Finally, Mr. Robert Blackburn of Chiron Corporation would provide the perspective of an in-house practitioner of patent law who has done both litigation and scholarly writing on patent law.

Dr. Goldstein reminded the audience of a recent Washington Post article about Craig Venter of Celera Genomics. Venter, according to the article, is filing more patent applications than he originally promised as his company sequences the human genome. Congress and other entities have criticized him for this. Although his company is filing thousands of patent applications, Dr. Venter, in defending Celera against critics, said that he anticipated his company having no more than a few hundred patents at the end of the day. Dr. Goldstein



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies Panel VI: Intellectual Property and the Public Domain: Sectoral Perspectives INTRODUCTION Jorge Goldstein Stern, Kessler, Goldstein, & Fox Mr. Goldstein opened the panel by commenting that patents reflect a “grand compromise” in giving inventors incentives to create new technologies—which it is hoped will widely benefit society—by allowing them to privatize the economic returns of an innovation’s benefits. There are questions about what we should privatize, for what length of time privatization should be ensured, and which institutions should benefit from privatization and which should not. The presentations in the panel would, Mr. Goldstein said, reflect the competing forces that run through the patent issue. Dr. Wes Cohen’s presentation would consider how different sectors view patents. Dr. Maryann Feldman’s presentation would focus on patents and universities, and the government-university relationship with respect to patents. Finally, Mr. Robert Blackburn of Chiron Corporation would provide the perspective of an in-house practitioner of patent law who has done both litigation and scholarly writing on patent law. Dr. Goldstein reminded the audience of a recent Washington Post article about Craig Venter of Celera Genomics. Venter, according to the article, is filing more patent applications than he originally promised as his company sequences the human genome. Congress and other entities have criticized him for this. Although his company is filing thousands of patent applications, Dr. Venter, in defending Celera against critics, said that he anticipated his company having no more than a few hundred patents at the end of the day. Dr. Goldstein

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies wondered whether Celera would in fact have just a few hundred patents by the time it finishes sequencing the human genome. In turning over the podium to the panelists, Mr. Goldstein said that the patent system reminded him of one of the animals from Dr. Dolittle, the Push-Me-Pull-You, the two-headed llama that wanted to go in two directions at once. The patent system, Dr. Goldstein said, may be a multi-headed llama, with a number of conflicting tensions in constant search of a grand compromise to determine in what direction it will finally go. SECTORAL VARIATIONS IN THE ROLE OF INTELLECTUAL PROPERTY Wesley Cohen Carnegie Mellon University Dr. Cohen began by saying that the title of the paper he was going to discuss is “Patents, Public Research, and Implications for Industrial Innovation in the Drug, Biotechnology, Semiconductor, and Computer Industries” and that his co-author is Dr. John Walsh of the University of Illinois at Chicago. The paper is part of a larger project being conducted in collaboration with Richard Nelson of Columbia University. Dr. Cohen said he would try to examine some of the assumptions underpinning the patent system, especially when it comes to the public-private interface and public research. By public research, Dr. Cohen said he meant research conducted by universities using public funds, and research conducted in federal laboratories. He also said that he would look closely at the assumptions underlying the Bayh-Dole Act. The rationale for Bayh-Dole was that there was an “urn full of untapped possibilities in universities and other public research institutions” and that universities needed the ability to patent this knowledge to fully exploit their R&D potential. This would serve as an incentive to universities to commercialize and provide the complementary R&D that would take these ideas into the commercial marketplace. Another part of this rationale was to give faculty members extra incentive to push their innovations from the university laboratory into the commercial arena. Dr. Cohen said that most of the circulation of university research is done through the traditional channel of publication in journals. The questions Dr. Cohen planned to address included the following: Do patents really protect industrial R&D? Does public research need to be privatized to ensure its commercialization? The industries to be examined are drugs, biotechnology, semiconductors, and computers. The data used in the analysis comes from the 1994 Carnegie

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies FIGURE 14 Importance of public research for industrial R&D. Mellon Survey of Industrial R&D. The respondents were R&D managers or unit heads in U.S. manufacturing industries. Overall, the survey received 1,478 responses, reflecting a 54 percent response rate. The respondent firms include the very large ones as well as those with just a handful of employees. For the firms Dr. Cohen discussed, the sample size was: drugs, 37; biotechnology, 21; computers, 34; and semiconductors, 24. One thing that Dr. Cohen would not discuss today, given the 1994 date of the collection of the data, is the growth of software patenting and the patenting of business methods, both of which are prominent topics today. As background, Dr. Cohen discussed the importance of public research to industrial R&D. The traditional academic wisdom on public research is that it is commercially important in a small handful of industries. In the full Carnegie Mellon sample, industrial research managers in all sectors reported that public R&D was important to their companies. For the four sectors analyzed in his paper, Dr. Cohen said that when asked whether research from universities or government labs suggested an R&D project or contributed to the completion of a project, the positive response was quite high. Figure 14 shows that for the category “Contributes to Project Completion” or “Suggests New R&D Project,” more than 50 percent of drug and biotech firms responded positively, while a

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies large portion of semiconductor firms did so. Only for the computer sector was the positive response relatively modest. These findings suggest the stakes involved with policies such as Bayh-Dole. If, for example, an industry reports that publicly funded R&D is of little commercial relevance, then policies such as Bayh-Dole may be of modest consequence. Dr. Cohen and his colleagues then asked industrial R&D managers about the channels through which they learned about public research. The channels of knowledge flow that managers were asked to consider were patents, publications, meetings or conferences, informal exchange, new hires, licenses, joint ventures, contracts, consulting, and temporary personnel exchanges. Across the entire sample of 1,478 firms, by far the most important channels were publications and meetings or conferences. Informal exchanges were also very important. In sum, the traditional channels of public science—publications and conferences—are the main sources of information for industrial R&D managers. For the four industries focused on for his research, Dr. Cohen said that papers and publications and meetings or conferences served as the primary means of communicating public research to industry. Looking more closely at drugs, biotech, computing, and semiconductors, Dr. Cohen said that “formal, market-mediated” channels, that is, “privatizable” mechanisms, are most important in drugs and biotech. Recalling an earlier figure, Dr. Cohen said that these two sectors are where public research is most important; in those industries, the public channels (i.e., papers and conferences) are most important. For semiconductors, the public channels, such as papers, are dominant, although informal exchanges and consulting arrangements are quite prominent in moving public research into the commercial arena. Although “privatizable” mechanisms play a role, Dr. Cohen said that given the weight placed by respondents on public channels, it is fair to conclude that it is not necessary for channels of information transfer to be private in order for public research to finds its way into commercial use. Even when considering the relative importance of private versus public channels, Dr. Cohen said that much of public research flows readily to the commercial sector using public channels. Dr. Cohen turned to the issue of how firms use patents to protect inventions, that is, the effectiveness of appropriability mechanisms for product innovations. Firms use a variety of ways beyond patenting to protect their innovations, such as secrecy, lead time, complementary sales and service arrangements, and complementary manufacturing. In the survey, respondents were asked to give the mean percentage of product innovations for which each mechanism was considered effective. The notion of effectiveness was defined as protecting the commercialization or licensing of an innovation that was protected using a particular mechanism, e.g., patenting or secrecy. As shown in Figure 15, there are differences across sectors in which mechanisms are more effective. In drugs and biotech, patents, secrecy, and lead time were the leading mechanisms. In semiconductors and computers, patents are

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies FIGURE 15 Effectiveness of appropriability for product innovations.

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies less important, with secrecy, lead time, and complementary sales and service being more useful mechanisms. Looking at patents alone, it is clear that biotechnology and drugs are two industries—either in comparison with semiconductors and computers or with all manufacturing sectors—in which patenting matters quite a bit. With the varying effectiveness of patents across sectors, one might ask why firms patent at all. In the case of biotechnology and drugs, Dr. Cohen said that indeed firms use patents for traditional reasons, namely to prevent other firms from gaining commercial advantage from their innovation. In semiconductors, where patenting is less effective, firms patent as part of a strategy to build a portfolio of knowledge in a particular industry to build “player strategy” in the field. Patents are also used in cross-licensing negotiations. In computers, standard licensing is more prevalent than in semiconductors, and patents are used in cross-licensing negotiations. In conclusion, Dr. Cohen said that most of the public research—that which takes place in universities or government laboratories—in the drugs, biotechnology, semiconductor, and computer industries is made available to these industries through public channels. This does not mean that privatization of some of these knowledge flows is a bad idea. These data suggest, however, that policymakers should carefully consider how essential privatization is in encouraging the commercial exploitation of public research. A key point is that importance of public research and the way in which it is transmitted to the commercial arena varies across industries. Therefore, what is at stake with respect to the Bayh-Dole Act also varies across industries. In biotech and drugs, Bayh-Dole may have the intended effect of promoting innovation by providing the legal framework for exclusive rights to intellectual property. But the importance of public channels in these industries (and public channels dominate private ones) suggests that exclusive rights are not critical to commercialization. Looking to the semiconductor industry, where patents are not terribly important, granting exclusive rights to public institutions for R&D is not likely to have much effect. Dr. Cohen sounded a cautionary note, saying that exclusive rights for public institutions may inhibit public-private partnerships. At Carnegie Mellon, Dr. Cohen recalled the story told to him by an electrical engineer, who said that he would gladly patent his research, but not press for license fees from corporate partners. If that happened, the corporate partners would not talk to him at all, fearing that each communication might result in a license fee or a legal dispute. The net result would be to inhibit the free flow of information between universities and industry. Questions From the Audience Dr. Cohen was asked whether the size of the firm affected the importance of patents. Dr. Cohen responded that he and his colleagues looked carefully at the

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies role of firm size, and found that large firms tend to view patents as being much more effective. The primary reason is that large firms have the legal resources to defend patents. Dr. Cohen added that the role of size with respect to patents varies across industries. In the biotechnology and drug sectors, small firms find patents to be very valuable; indeed, patents are absolutely essential to obtaining venture capital financing. POST-BAYH-DOLE UNIVERSITY-INDUSTRY RELATIONSHIPS Maryann Feldman Johns Hopkins University In talking about university-industry relationships after passage of the Bayh-Dole Act, Dr. Feldman said that she would synthesize a multi-year research project funded by the Andrew Mellon Foundation. These studies have explored universities’ responses to the Bayh-Dole Act. David Mowery of Berkeley and Richard Nelson of Columbia conducted the first set of Mellon studies on this topic; they examined the University of California at Berkeley, Stanford University, and Columbia University. The second set has looked at Johns Hopkins University, Duke University, and Penn State. The objective is to increase the number of university intellectual property databases that are available for analysis. The studies also aim to study the relationship between industry and universities and understand that relationship in an evolutionary framework. Dr. Feldman said her presentation would attempt to accomplish the following: Interpret some of the changes in university-industry relationships post-Bayh-Dole; Demonstrate the variations in which universities have responded to the new opportunities presented by Bayh-Dole; and Offer an interpretative framework to understand the evolving industry-university relationship. The post-Bayh-Dole era for the university-industry relationship has been marked not just by a change in the intellectual property regime, but also by a number of other changes in the environment. In many sectors, such as biotechnology and computing, universities have been sources of “high-opportunity” technologies over the past 15 to 20 years, and industry has sought access to these technologies. The new technological opportunities available from universities have been accompanied by a growing need for universities to find new sources of research funding. Competition from researchers has also grown; more researchers are out there seeking grants to conduct research.

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies Although Bayh-Dole’s passage in 1982 serves as an important policy water-mark, the six universities under study were patenting prior to that and, since its passage, in a changing environment. Nonetheless, Bayh-Dole deepened universities’ existing patenting activities, while also expanding patenting to universities that were inactive in patenting in the past. Under Bayh-Dole, all universities acquired broader intellectual property rights, but the Mellon studies found great diversity across universities with respect to institutional policies. For example, universities place different weights on how founding a company is credited; some may look very positively on this when evaluating professors’ performance, others may place little weight on it. Different universities may look differently on a professor funded by industry versus one being funded by the National Science Foundation or the National Institutes of Health. In many universities, commercial activities are still regarded as “second tier” sorts of undertakings. Universities also vary in the amount of funding delay they will tolerate as a consequence of working with industry. Finally, the amount of authority—in terms of funding and charters—accorded to university technology transfer offices varies tremendously across universities. Many of these variations simply reflect the different cultures in the universities studied. With respect to research results, Dr. Feldman displayed a table showing total research expenditures among selected universities and research royalties (Table 3). As the table shows, there is not a perfect match between overall university research expenditures and royalties received. The University of California system spent close to $1.6 billion in 1997 on research and also received the most dollars back in royalties, $61 million. The second largest royalty generator, however, was Columbia University, which received $46 million in royalty revenue, and it was last in total research expenditures among universities listed. Johns Hopkins University, in contrast, spent close to $1 billion on research in 1997, but had only a meager $4.6 million in royalty revenues. Exploring the data further, Dr. Feldman displayed a table (Table 4) showing measures of the productivity of university research spending, using invention disclosures and licensing revenues as metrics. For a university’s invention ratio—the number of invention disclosures per $1 million in research expenditures—Stanford, Wisconsin, the Massachusetts Institute of Technology, and Columbia had the highest ratios. In terms of royalty per license, Columbia topped the list, followed by Stanford and Wisconsin. The patents and licensing revenues on which the economics literature has focused with respect to innovation give only a very narrow snapshot of what has been going on with universities in the post Bayh-Dole era. Universities have been undergoing a search process as they try to adapt to Bayh-Dole. Should universities retain their traditional role as generators and disseminators of knowledge? Or is revenue generation now a legitimate goal? Dr. Feldman and her colleagues have been conducting a series of interviews on these topics at universities, and they have discovered a great deal of conflict

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies TABLE 3 Differences in University Technology Transfer Activities University Total Research Expenditures Adjusted Royalties Received Licenses Generating Royalties Start-ups Formed UC System $1,586,533,000 $612,800,000 528 13 Johns Hopkins $942,439,696 $4,686,519 103 3 MIT $713,600,000 $19,860,549 255 17 WRF $528,602,441 $11,478,605 142 25 Michigan $458,500,000 $1,708,939 83 6 Stanford $391,141,224 $34,014,090 272 15 WARF $379,600,000 $17,172,808 133 2 Harvard $366,710,262 $13,402,273 232 1 North Carolina $263,517,405 $1,665,909 61 2 Columbia $244,100,000 $46,105,192 201 4   SOURCE: Association of University Technology Managers, 1997 on these questions. In the years immediately after Bayh-Dole, universities often granted exclusive licenses to industry for their innovations; now universities are quite selective when it comes to conferring exclusive licenses. Universities have also developed new mechanisms for commercialization; in 1997, there were 248 university-generated start-up companies, and universities took an equity position in 70 percent of them. In thinking through a framework for understanding industry-university relationships, Dr. Feldman and her colleagues have found that commercializing university technology will involve multiple types of transactions. Multiple licenses for a technology or family of technologies are one approach that universities may employ. In exchange for a license from a university, companies will often sponsor a research project at the university. This helps the company build exper- TABLE 4 Productivity of University Research Spending University Invention Ratio Royalty Per License UC System 4.51 $116,061 Johns Hopkins 2.43 $45,500 MIT 5.04 $77,885 WRF 5.30 $80,835 Michigan 3.66 $20,590 Stanford 6.34 $125,052 WARF 5.24 $129,119 Harvard 3.25 $57,768 North Carolina 3.57 $27,310 Columbia 6.02 $229,379

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies tise in a technology; a company may also do this through hiring students that also work with the faculty principally involved with the technology. Dr. Feldman and her colleague have also observed companies making gifts and endowments to universities as the relationship between a company and a university has developed. An important point to keep in mind, Dr. Feldman said, is that the technology transfer transaction is not the most important issue, but rather it is the relationship between a university and private sector partners that develops over time. Understanding those relationships is the best way to appreciate the evolution of university-industry ties in the post-Bayh-Dole era. Universities and companies have, therefore, engaged in a very active exchange of information, using transactions such as licenses, sponsored research, consulting arrangements, recruitment, equity arrangements, support for start-ups, and others. Precisely what transactions dominate will vary depending on the sector, the technology, and the size of the company, as Dr. Cohen said in his presentation. The formal and informal rules and norms of universities will also shape the evolution of their relationships with companies. In terms of exogenous policy factors that will influence these interactions, Dr. Feldman pointed to the intellectual property regime and funding availability. Dr. Feldman and her colleagues have started interviews to develop case studies of particular firms, universities, and technology areas to gain a greater sense of the factors that affect the evolution of university-industry relations in the post-Bayh-Dole era. In conducting archival research on the history of patenting at Johns Hopkins University, Dr. Feldman uncovered correspondence between the National Research Council and the dean of the medical school from 1949. At the time, NRC was conducting a survey of university intellectual property policy, and it was inquiring about the possible participation of personnel from Johns Hopkins in a conference on the topic. The response from the dean of the medical school stated that individuals in the school would neither be interested in patenting innovations that may impact the public health nor participating in the NRC conference. In fact, the medical school viewed it as “undesirable” for doctors associated with Hopkins to patent their inventions or discoveries as they pertain to public health. As of 1949, no member of Hopkins’ medical school faculty had applied for a patent. Based on her interviews, Dr. Feldman said that Johns Hopkins has a legacy of discouraging patenting. Faculty members hired in the 1970s, coming from places such as Stanford that encouraged patenting, were concerned about a culture at Hopkins that discouraged such activity. In recent years, Johns Hopkins University has been trying to change this culture.

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies Conclusions Dr. Feldman raised the following issues in conclusion: What are the most effective means to organize and conduct technology creation and diffusion? Are the financial and transaction costs of doing science increasing in the post Bayh-Dole era? Does emphasis on patenting create divided loyalties for faculty and does this hurt students? Questions From the Audience Dr. Kathy Behrens asked whether Dr. Feldman and her colleagues looked at a) the number of inventions that created the royalties at universities and b) the year in which the intellectual property was created that generated the revenue shown for 1997. Dr. Feldman responded that the data does not yet exist that permit these important questions to be answered. One of the motivations of the Mellon studies is to gain access to offices within universities that will enable the necessary data to be assembled. From her research, Dr. Feldman said that it appears that certain licenses generate a great deal of revenue and that there is a gestation period until licenses begin to generate returns. It seems clear, said Dr. Feldman, that there are just “a few big hits” in the university-patenting world. Mr. Goldstein asked whether Dr. Feldman’s research is consistent with Dr. Cohen’s that licenses in the biotechnology and drug area are more valuable than those in engineering or electronics. Dr. Feldman responded affirmatively, noting that at Johns Hopkins approximately 90 percent of the university’s intellectual property portfolio is in the biosciences. That figure is very similar for Duke University and Penn State University. At Johns Hopkins, Dr. Feldman continued, the largest revenue source from a license is for software from the School of Public Health; this raises classification issues for license revenues. INTELLECTUAL PROPERTY AND BIOTECHNOLOGY Robert Blackburn Chiron Corporation Mr. Blackburn began by saying that the number of biological and medical threats that we face today is growing, if we take seriously what epidemiologists tell us about the advance of HIV in the Third World and the growth of new drugresistant strains of tuberculosis. In a very real sense, Mr. Blackburn said, we are in a biological war. The U.S. patent system is designed to encourage progress in

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies the “useful arts and sciences,” and the need for such encouragement has never been so urgent. For the biotechnology industry, the patent system has been superb at protecting products sold in the marketplace and methods of manufacturing. Where there has been significant controversy is in the area of research tools. Research tools, in the eyes of some industry observers, should not be patentable. Others believe that they should be patentable because research tools represent something of value. Mr. Blackburn said that this proposition should be closely scrutinized, and he suggested that it might not be true. If it is not true, the concern arises that venture capitalists may not have sufficient incentives to invest in such tools. The debate over the patentability of research tools is complex. It includes issues such as the statute of limitations, the clinical trial exemption under the Hatch-Waxman Act, and reach-through royalties (i.e., whether they can be awarded by a court). If one is not cognizant of these complexities, then one is debating with only 10 percent of the relevant information at hand. Mr. Blackburn said he would discuss these complicating issues, with the hope of moving beyond some of the more emotional elements that sometime arise in debates about the biotechnology industry. The Evolution of Drug Discovery Mr. Blackburn first discussed some of the differences between the traditional pharmaceutical industry and the biotechnology industry. The process of discovery in the traditional pharmaceuticals industry involves finding a lead chemical compound that has a desirable biological property and that can be administered orally. Often these chemicals were found by happenstance, and the real science begins when a pharmaceutical chemist optimizes it into a compound that has less toxicity, more activity, and is more biologically available. This is often a slow process, as each version of the drug has to be synthesized and tested. Biotechnology is very different, in that it involves finding a disease with a large protein and then finding the gene for that protein. The next challenge is developing a way to make the protein in a reasonably cheap manner by recombinant DNA expression. These proteins, however, are generally expensive to make, although it is less expensive to make them using recombinant DNA than by finding them in nature. Proteins made by recombinant DNA are also not usually orally available, so there are limitations to this approach. Examples of these types of inventions are Factor Eight for hemophilia, Interlukin 2 for cancer treatment (which also has promise for AIDS), a hepatitis B vaccine that is a recombinant yeast expression of the vaccine, and human growth hormone. All of these are large proteins made by recombinant DNA technology. Today’s multidisciplinary biopharmaceutical research has brought the tools of chemistry and biotechnology together. This interaction is driving modern

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies pharmaceutical science into an era of unprecedented drug development, and it represents nothing short of a revolution. A technique called combinatorial chemistry, for example, allows a chemist to make a library—essentially a test tube—with 100,000 randomly generated chemical structures. This is in contrast to traditional methods that require synthesis of individual chemical compounds. With the invention of targets, which is the site of action for particular drugs, chemists test the randomly generated compounds against the targets. Other scientists have invented high throughput assays that automate this process. These targets are often proteins that have been discovered using recombinant DNA techniques or genes themselves. When an individual compound is found in a library, it is known as a “hit,” but scientists will still not know what it is, as it is in the complex mixture. So-called “deconvolution technology” pulls the compound out of the mixture and gives scientists its structure. This allows them to go back to the laboratory to make the compound in large quantities. This amalgam of technologies can create a number of lead compounds in a short period of time. From that point, more traditional pharmaceutical chemistry, such as optimization, can take over, as well as newer tools of “rational drug design.” The power of this technology is truly remarkable. For diseases for which no drug therapy had been found for decades, scientist are finding multiple lead compounds, sometimes in a matter of weeks. In the coming decades, we will have, in all likelihood, multiple drugs for diseases for which we have had either poor treatments or no treatments. New Developments in Biotechnology and Patenting There is, however, a problem with this amalgam of technology. The current patent system may not provide the right balance of protection and freedom of operation that will allow these technologies to flourish. These technologies are not the product being sold in the marketplace; they are used in the research phase of drug development. The patent laws we have today are superb at protecting final products, but are untested when it comes to patents covering research tools. The problem is very significant because it relates to the most critical technology that we have to benefit human health. To understand fully why patent law has been untested on research tools, it is necessary to understand what the scope of a patent is and what a patent covers. If you invent a research tool, the Patent Office does not grant a patent to you on the things that people invent using the tool. The inventor receives a patent only on the research tool, and perhaps the method of using the tool, as well as the reagents and machinery involved with the tool’s use. The patent will not cover the downstream product. Possible infringement of the patent will occur in the research phase, prior to a drug entering the marketplace. Detecting infringement will therefore be very

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies difficult, because companies do not usually publish their lead compounds or optimization research used in drug development. This is where secrecy enters the picture. Recalling Dr. Cohen’s finding that secrecy is important to drug companies, Mr. Blackburn said that keeping research methods secret is key to pharmaceutical companies having quick time-to-market. By the time a patent-holder discovers that a research tool has been improperly used, one usual remedy, an injunction against use, is ineffective. Granting damages is another possible remedy, but there is a six-year statute of limitations in the United States on filing an infringement suit. However, because of the time it takes to develop drugs, it would not be unusual for the infringement to be discovered after the statute of limitations has expired. The Hatch-Waxman Act Even if a holder of a research tool patent clears these hurdles in infringement litigation, the Hatch-Waxman Act presents another barrier. This legislation was enacted to permit generic drug companies to run clinical trials on a patented drug before the patent on that drug expires so that generic drug companies would be ready to enter the market as soon as the patent expired. A court had found that running a clinical trial was a patent infringement, and Congress overturned the ruling on the rationale that the court ruling amounted to an unwarranted patent extension. Mr. Blackburn said the act exempted “from infringement those activities reasonably related to submitting data to the FDA [Food and Drug Administration].” The Act does not limit the types of patents to which the legislation could be applied. Because the Hatch-Waxman Act was passed before the invention of research tools, Congress passed the law only with patents for drugs in the marketplace in mind. It was impossible to consider the act’s effect on research tools. Parties that have allegedly infringed research tool patents have used the act as a defense, arguing that Hatch-Waxman exempts them from infringement suits. If that argument stands, the end result is that it may be impossible to infringe patents on research tools. Recent Court Decisions Assuming further that a research tool patent holder has cleared the HatchWaxman hurdle and is in court, problems arise from recent court decisions that apply patent law doctrine developed for the synthetic chemicals industry to biotechnology. Patents in biotechnology typically involve discoveries of things found in nature—a DNA sequence for example—which are subject to a wide degree of variation in specific applications, and the variations are relatively obvious. In applying patent law developed for the synthetic chemical industry, courts have rightly recognized as a significant invention the compound that will

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies be sold in the market. But in controversial cases, such as The University of California v. Lilly, courts have extended this doctrine, and this has resulted in a significant reduction in the scope of patents for biotechnology inventions. Mr. Blackburn said that the doctrine made sense when the invention was making the new structure, as in synthetic chemicals, where the structures simply did not exist before. It probably does not make sense, argued Mr. Blackburn, to apply that doctrine when the invention involves finding preexisting structures in nature and making that information available so that it can be turned into products. Those in favor of the Lilly decision might argue that broad patents might be just as harmful to innovation in the biotechnology industry as the unavailability of patents. Mr. Blackburn said he did not envy the Federal Circuit Court that would have to ultimately strike the right balance, and he pointed out that the courts are the only governmental bodies presently addressing these basic policy issues. A problem is that the legal process limits the amount of input the court considers. In the Lilly case, the court had to address about a dozen technical issues, but the petitioners had to limit their briefs to no more than 50 pages. That is a small information base with which to adjudicate very technical scientific and legal issues. Delay is another problem. In the University of California v. Lilly case, this decision—which now serves as the guide on how to make a patent application in biotechnology—was handed down 20 years after the invention in dispute was made. Two decades is simply too slow in an industry as dynamic as biotechnology. On average, it takes 10 to 15 years from “benchtop to marketplace” for drugs, and under the U.S. patent system, it is very difficult to challenge patents prior to the date of award. A Scenario of New Drug Development To elaborate on this point, Mr. Blackburn presented a hypothetical situation about the development of a drug in a biopharmaceutical company. Suppose you are the director of research at this biopharmaceutical company. You ask the firm’s patent attorney whether a particular patent will block the development of a drug you would like to develop. The patent in question covers a final product, and your company’s patent attorney says that you will probably infringe upon the patent when the drug comes on the market. In the meantime, however, Hatch-Waxman will probably protect research you conduct. The patent attorney adds that, in her opinion, the patent is invalid; it involves a doctrine that the courts have yet to address, and in her reading of current trends, she estimates a 60 percent chance that the patent would be defeated in federal circuit court—in 15 years. As research director, you then turn to your business development people, who cannot procure a license to the technology on reasonable terms. The 60

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies percent chance that the patent will be thrown out is causing you great pause. The chances of any one drug making it through to the market are very low; to further discount the chances by 40 percent makes it an unacceptable expenditure of money. The cost of a successful research program to develop the drug is likely to approach $200 million. In this case, a 40 percent chance exists that a product could not be sold. However, the potential upside of the drug is so high, several hundred thousand dollars in research expenditures today seems worthwhile to test the patent. To obtain a declaratory judgment that the patent is invalid, however, requires a case and controversy; in this case, this means that a reasonable chance exists that the patent holder will litigate. The patent holder, not wanting a court to possibly invalidate his patent, declines to provide a letter or any evidence to suggest that litigation is likely. Having a patent declared invalid, your patent attorney informs you, does not invoke the already weak reexamination procedures available from the Patent Office. A federal court would really have to make a ruling. That sort of litigation would cost approximately $5 million, which is far more than the several hundred thousand dollars that you wanted to invest in exploratory research. In the end you, as research director, kill the project, even though further development seems technically feasible and promising, and there is a 60 percent chance that the existing patent would not block you. Biopharmaceutical companies face this sort of decision every day. Some attorneys might argue, said Mr. Blackburn, that this is how our patent system works and it has worked well for several centuries. While this may be a good patent system, argued Mr. Blackburn, “it is a lousy industrial policy.” Patenting Policy and the Climate for Innovation One possible solution to these problems is to discontinue patenting research tools. This, of course, lessens incentives to innovation in this field. What is needed, Mr. Blackburn argued, is a system in which drug developers can obtain reasonable certainty about the climate for innovation in the industry. Such certainty must be attainable at a low cost and in a timeframe so that investment decisions are not unduly hindered. Legislation could perhaps address some of these problems. For instance, legislation could make it easier to obtain a declaratory judgment by defining a “case and controversy” to include the existence of a patent that might block the development of a new drug. Legislation could also address the current application of chemical patent doctrine—developed over 50 years ago—to biotechnology. Biotechnology is, after all, based on completely new science, and thus calls for new rules. The Japanese and German models of patent enforcement, said Mr. Blackburn, should be closely considered as solutions to U.S. problems. In those countries, expert tribunals within their patent offices hear all challenges to patent

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies validity. This is much cheaper and faster than federal court litigation in the United States. Federal courts would still hear cases on infringement and damages, which are reasonably predictable. Mr. Blackburn did not advocate wholesale adoption of the German or Japanese systems, because in some respects they run counter to U.S. notions of due process. But those systems offer useful guidance in developing alternatives to an American system that currently restricts the subject matter on which patents could be granted. The stakes are important, Mr. Blackburn concluded, because “getting it [the patent regime] wrong” could severely suppress the entrepreneurial spirit that drives innovation in biotechnology. In today’s highly competitive climate, “the inventions you don’t make could kill you.” DISCUSSION Mr. Goldstein commented that he sees two important issues coming from the presentations. With respect to the papers by Dr. Cohen and Dr. Feldman, he wondered whether there were differences in the computer industry between hardware and software in terms of the importance of patents. Given the changes in the past five years in the legal regime governing software patents, Mr. Goldstein said that software patents are likely to become much more important. Labeling them as “computer patents” may thus be somewhat misleading. Mr. Goldstein suggested a fertile area of research would be exploring how legal decisions affirming software’s patentability would affect industry behavior. Mr. Blackburn’s discussion of research tools brings a number of important themes together. Research tools, said Mr. Goldstein, will not only be one of the great high-technology developments in finding new drugs in the next 10 to 15 years, but will also cause the most friction between industry and universities. In Mr. Goldstein’s law practice, he has had more problems with either academics or industrial researchers being irritated with having requested a research tool, and receiving a lengthy letter from a lawyer with terms about “reach throughs,” future intellectual property rights, and other legal conditions. The NIH has issued guidelines on this, but the legal problems surrounding research tools are greatly straining relations between universities and industry. Dr. Stephen Merrill asked Mr. Goldstein to comment further on his earlier statement about patents as a public source of technical information. Mr. Goldstein responded by saying that it is clearly in the best interests of everyone who files for patent applications to know what the best “prior art” is. It does no one any good to receive patents that are invalid; they are a real burden on the system. Cynics might say that an invalid patent in hand is better than no patent, but in Mr. Goldstein’s experience, he does not see applicants “hiding their head in the sand” by ignoring prior art in order to obtain a patent at all costs that will then “hold up the industry.” In the software area, critics say that the Patent and Trademark Office issues

OCR for page 173
Capitalizing on New Needs and New Opportunities: Government-Industry Partnerships in Biotechnology and Information Technologies many invalid patents in software; indeed, Mr. Goldstein said that some software patents may be declared invalid for “obviousness reasons.” The Patent Office lacks the manpower and databases to understand the range of products and innovations in software. However, it is not an industry-wide phenomenon where software firms file for patents, which they suspect are invalid, as a strategy to hold back technical progress by other firms.