College of American Pathologists
Printable Version

  Friend or foe? Looking at gene patents


CAP Today




February 2010
Feature Story

William Check, PhD

What’s the difference between a human gene and an adjustable brake pedal? Is a human gene different from a computerized trading program? In the eyes of patent law, these items are all essentially equivalent, at least insofar as patents on them are governed by the same criteria. And that’s why patent lawyers, members of the biotechnology industry, and savvy molecular pathologists are paying attention to two Supreme Court cases, one decided in 2007 (KSR v. Teleflex) and one pending (Bilski v. Kappos).

In KSR, the High Court ruled that the Court of Appeals for the Federal Circuit had used overly restrictive criteria in determining that an adjustable brake pedal combined with an electronic sensor was not an obvious invention. With this decision, the Supreme Court “liberalized the criteria for obviousness, requiring a more flexible approach than that taken by the Federal Circuit,” says Roger D. Klein, MD, JD, medical director of molecular oncology at BloodCenter of Wisconsin and clinical assistant professor of pathology at Medical College of Wisconsin. The Supreme Court ruled that “obviousness to try” could render a patent obvious where there is a demonstrated need for a discovery and a finite number of identified, predictable solutions.

“The implications for many DNA patents are clear,” says Dr. Klein, “because many, if not most, patented genes were mapped to a chromosomal region prior to their discovery. Moreover, many are involved in biochemical pathways that have been previously implicated in specific diseases.”

In Bilski, the court is evaluating a challenge to methods patents. Hans Sauer, PhD, JD, associate general counsel for intellectual property at the Biotechnology Industry Organization, says the case involves a method of hedging risk in energy commodities trading, making it reasonable to ask why the biotech industry is paying attention. “The Patent and Trademark Office has problems examining business methods patents, which is why the issue is important,” Dr. Sauer says. The patentability of a relationship between a biomarker and a disease process or response to therapy—such as measuring EGFR variants in lung cancer—could be affected by the Bilski decision. “An unprecedented number of amicus briefs were filed on Bilski—more than 60,” Dr. Sauer says.

Welcome to the wonderful new world of gene patenting. For 30 years Congress, the courts, and the U.S. Patent and Trademark Office encouraged and facilitated gene patenting. More recently there has been a push to reconsider lenient interpretations of patentability. The lawsuit brought by the American Civil Liberties Union against Myriad Genetics is an attempt to change the rules by court action. A case for sum­mary judgment was heard Feb. 2; though the judge hadn’t yet ruled at press time, it appeared the case would go to trial. Action by Congress is advocated in a report drafted by the Task Force on Gene Patents and Licensing Practices of the Secretary’s Advisory Committee on Genetics, Health, and Society and adopted by SACGHS on Feb. 5.

Of course, the “biotechnology industry” is heterogeneous, and people from its different sectors have different views of whether genes should be patented and licensed. Directors of molecular labs in hospitals or medical centers typically take a dim view of patenting and exclusive licensing of genes, though what they consider allowable or reasonable varies.

“Human genes should not be patentable, following the argument that they exist in nature and are not sufficiently manmade inventions to justify patents,” says David H. Ledbetter, PhD, director of the Division of Medical Genetics in the Department of Human Genetics, Emory University School of Medicine. “For me as a genetic diagnostics person, the practical problem is when we have a gene testing monopoly, where only one lab is allowed to offer that test. That is analogous to having certain medical treatments available only in one medical center.”

Since patents are currently legal, Dr. Ledbetter suggests flexible licensing to encourage the broadest access to genetic technology. “Most gene discoveries and gene patents occur at academic institutions by investigators using federal funding,” he says. “If individual scientists and their institutions would follow existing guidelines recommending nonexclusive licensing of genomic patents, there would be less of a problem.”

To Debra G.B. Leonard, MD, PhD, professor and vice chair of laboratory medicine at Weill Cornell Medical College, having a sole provider of a medical service is not in the best interests of the public health. Restricted access, less incentive to improve tests, and loss of competitive pricing are some of the harms. “My concern has been and continues to be people who use a gene patent as a way to control an area of medical practice.”

One of the things that Karen Mann, MD, PhD, associate professor and director of molecular hematopathology at Emory, considers important about exclusive licensing is that the patient can’t get a laboratory second opinion. “You need more than one lab doing a test so you can have proficiency testing. There are alternative types of proficiency testing, but they are not as good as comparing to peers,” she says.

“Patenting of human genes per se is foolish and a waste of everyone’s money,” says Sherri J. Bale, PhD, FACMG, co-president and clinical director of GeneDx, Gaithersburg, Md. “If licensing is appropriate and loose, there is no real harm done. But exclusive licensing on patented genes—that is an absolute disaster. It impairs our ability to provide good clinical care.” In Dr. Bale’s view, current policies of the patent office and of many university technology transfer departments benefit investors much more than patients. (See “A case of limited clinical access”)

Arupa Ganguly, PhD, associate professor of genetics and director of the genetic diagnostic laboratory at the Hospital of the University of Pennsylvania, notes that exclusive licensing restricts patients’ choices. “People should have the option of going to an academic center as opposed to a commercial lab,” she says. Dr. Ganguly’s concern reflects more than a turf battle. Before she was enjoined from doing BRCA testing, she had an unusual experience with the family of a patient who had a mutation in the BRCA1 gene. “We could not find the BRCA1 mutation in family members,” she explains. “Yet the disease presentation suggested a genetic predisposition.” Dr. Ganguly’s laboratory sequenced the BRCA2 gene and found that it also contained a mutation. “This was a rare case, where the same family had mutations in both BRCA genes,” she says. “This could only be discovered in an academic setting. It would not have been done in a commercial lab.”

Exclusive licenses not only clear the market of existing tests, they have a chilling effect on laboratories that might want to develop or offer a test. “We have a lab-developed test for JAK2 but are hesitating about offering it because of the probability that the patent holder will make us stop,” says Jan A. Nowak, PhD, MD, director of the molecular diagnostics laboratory in the Department of Pathology and Laboratory Medicine at Evanston (Ill.) Hospital, NorthShore University HealthSystem. A particular problem is that Ipsogen’s test kit for JAK2 is still research use only. “Until their kit goes through FDA, there is no assay that can be used clinically,” Dr. Nowak says. “We could validate it ourselves, but how would we do that without sequence information?”

Dr. Klein of the BloodCenter of Wisconsin raises a concern about a type of patent that is more complex than those on sequences or specific mutations—genetic association patents, which are the genetic equivalent of the “methods” patents that are the focus of the Bilski case. “To me what is important now, and what is happening on a routine basis, is the patenting of correlations between gene variants and either diagnosis, prognosis, or responsiveness to therapy,” Dr. Klein says. “That latter application gets to the heart of personalized medicine.” One of the main targets in the suit against Myriad—the claim for correlations between mutations in BRCA genes and susceptibility to breast and ovarian cancer—is an association patent.

Contrary opinions prevail among industry proponents. “The real issue is whether genes should be patentable,” says Dr. Sauer of the Biotechnology Industry Organization. “The only way to answer that is through Congress and public debate. Courts can’t decide that.” Speaking on behalf of BIO, he says: “As a matter of patent law we do think that isolated and purified DNA is patentable. We also think there are policy reasons why genes should be patentable. The same patents that protect genetic diagnostic tests also protect therapeutic products developed under that technology. Much different levels of investment are needed for diagnostics as opposed to therapeutics.”

Timothy T. Stenzel, MD, PhD, chief scientific officer of Quidel Corp., warns that any revision in current U.S. patent law will severely disrupt the funding and activity of the biotech industry. “Gene patents stimulate gene discovery, produce greater investment in research, and promote development of new diagnostics and drugs. I am broadly supportive of current U.S. patent law. It is vital for the future benefit of patients.”

Even within the biotechnology industry, however, dissent exists, particularly in the “tools” sector, such as companies that make gene chips. Barbara Caulfield, former general counsel for Affymetrix and now managing partner of the Silicon Valley office of Dewey & LeBoeuf, says: “Companies in the research tools sector of the biotechnology industry aim to put as many genetic signatures on a chip as possible so their clients in universities and genetic research institutions can conduct the broadest possible research. If a company claims to own a naturally occurring gene, that impedes the ability of tools companies to make the most useful products.” That’s one reason why tools companies have been in the forefront in challenging gene patents, according to Caulfield. A second, more fundamental reason: “Gene patents are not authorized by patent law.” Current patent office practice has never been challenged directly in court, Caulfield notes. She considers the ACLU suit against Myriad Genetics a good step in that direction.

As for the warning that any change to patent law will disrupt the biotechnology industry, Caulfield counters, “Gene patents disrupt the biotechnology industry.”

Holding the middle ground among biotechnology companies is Celera Corp. “We think patents on genes are fine,” says Tom White, PhD, Celera’s chief scientific officer. “It’s not gene patents per se that cause problems, but licensing practices.” Dr. White suggests adjusting licensing practices, such as requiring recipients of government grants to issue nonexclusive licenses. Celera has identified and patented a genetic marker called KIF6 that signals elevated heart disease risk and response to statin therapy. It has granted four nonexclusive licenses on KIF6 to date.

Dr. White warns against “unintended consequences” of the lawsuit against Myriad. While the suit focuses on diagnostic applications of gene patents, he says, “They may not quite realize that if their lawsuit succeeds, gene patents will be outlawed not only for diagnostics but also for gene-based therapeutics, which have had a profound impact on human health.”

Jeffrey A. Kant, MD, PhD, wonders if it is possible to debate this issue rationally. “Personally, I don’t accept the argument that patents are essential for economic success,” says Dr. Kant, professor of pathology and human genetics and director of the Division of Molecular Diagnostics, University of Pittsburgh Medical Center. “The argument from the enterprise side is that without patents there will be no innovation. I’m not sure there is any evidence that is true, and there is some evidence it may not be true.” But this line of thinking, he says, is such a fundamental part of the system that it is “difficult to get people to question it.”

Arthur L. Caplan, PhD, director of the Center for Bioethics, University of Pennsylvania, is not connected to either side of the gene patent debate. In a commentary on the ACLU suit, Dr. Caplan wrote, “Tossing out the Myriad patents would indeed imperil thousands of other patents, and courts tend not to want to cause that much turmoil” ( Asked about this view, he told CAP TODAY: “I don’t think the courts would be willing to re-examine the entire patent system just because Myriad has acted unfairly. It would cause too big an upheaval in the system.” However, he added, “I think patenting of genes has gotten way out of control. There is no inviolate principle that patents cannot be changed, withdrawn, or altered. A patent is a privilege, and society could take it away. I don’t think courts should do that, but they could raise the bar higher for future patents while grandfathering current ones.”

Dr. Caplan put forth in 1998 three ethical arguments against gene patents. For example: “[P]atenting is morally wrong because it allows individual ownership over something that is community property or a public good” ( While acknowledging that ethics is not at the heart of the lawsuit, he told CAP TODAY, “I still hold to the ideas in that article.”

A timely workshop on gene patents was held at the 2009 meeting of the Association for Molecular Pathology. Members of the AMP have a special interest in this topic, because the AMP is the lead plaintiff in the suit against Myriad. (The ACLU is technically plaintiff’s attorney.) On May 12, 2009 Dr. Nowak, who was then president of the AMP, announced in an email to the membership the decision to accept the ACLU’s invitation to be lead plaintiff. “At each stage [of the decision process] the voting left no doubt that AMP Leadership favored joining the suit,” he wrote. Other plaintiffs include the College of American Pathologists, American Society for Clinical Pathology, American Medical Association, American College of Medical Genetics, and several molecular pathologists and patients ( The lead defendant is the U.S. Patent and Trademark Office, or USPTO.

“Clearly, gene patents are very much in the news and of interest to our members,” says Emory’s Dr. Mann, who chaired the AMP workshop. “One big reason is the lawsuit. But I think it’s more than that. For instance, SACGHS just came out with a document about gene patents. In addition, some of our members are quite sophisticated about this topic while others knew less.” The AMP workshop, she adds, was intended in part to get everyone to a level playing field.

At the workshop Dr. Klein reviewed the legal basis of gene patents. Article I, section 8 of the Constitution says that patents are intended “to promote the progress of science and the useful arts.” “The Constitution itself granted Congress the right to create laws for intellectual property rights,” Dr. Klein said. The last full codification of patent law was the Patent Act of 1952. In 1982 the Court of Appeals for the Federal Circuit (CAFC) was created to facilitate uniformity in patent case law.

It’s no coincidence that the CAFC was created two years after the first court case involving a genetic patent, Diamond v. Chakrabarty, about a microorganism bioengineered to digest oil spills. Initially, the USPTO rejected the application, but the Supreme Court reversed, quoting the famous phrase from the 1952 Patent Act that patents could be granted on “anything under the sun that is made by man,” which, the court said, “could include living organisms.”

It is a long way from patenting a recombinant organism to patenting the sequence of an isolated gene. To make this journey, the USPTO and the CAFC traveled in the time-honored vehicle of precedent. “Historically, there is a precedent for patenting isolated purified chemicals used as therapeutics, even from the human body,” Dr. Klein said. Examples include epinephrine and prostaglandins. “Now this precedent has been applied to DNA as a chemical.”

In a CAP TODAY interview, Dr. Klein critiqued this interpretation. “Of course, DNA is a chemical, but it is also much more. Its primary function is to encode specific information for functional proteins. In most genetic testing, we are concerned with the sequence of DNA code that transmits information into a sequence of amino acids in proteins. If, on the other hand, you cloned DNA and used it to manufacture a drug, one could make a fair argument that it is similar to other chemicals.” Because of this distinction, the AMP/ACLU lawsuit and the SACGHS Task Force on Gene Patents and Licensing Practices focus on the use of genes for diagnosis.

A major challenge to current interpretations of patent law was posed by a 2006 case, LabCorp v. Metabolite Laboratories, which involved a patent on the relationship between elevations in blood levels of the amino acid homocysteine and deficiencies in vitamin B12 and folate, and which Dr. Klein has analyzed (Klein RD, Mahoney MJ. J Law Med Ethics. 2008;36:141–149). Apparently for technical reasons, the Supreme Court declined to take the case. However, a dissent written by justice Breyer and joined by justices Souter and Stevens expressed doubt about biological association patents. “At most respondents have simply described the natural law at issue in the abstract language of a process,” justice Breyer wrote. Dr. Klein says, “Patents on correlations or associations between genetic variants and the diagnosis, prognosis, or prediction of response to therapy are not true methods but instead use artful patent drafting techniques to claim the underlying natural or biologic phenomenon.”

“People look at genes as simply pieces of DNA, as molecules,” says Dr. Leonard of Weill Cornell. “But in reality DNA is an information system, the information system of our body.”

In her talk at the AMP workshop, Dr. Leonard cited legislative attempts to alter gene patenting rules. Rep. Lynn Rivers of Michigan introduced a bill in March 2002 to exempt from patent infringement claims clinicians and researchers who use gene-based diagnostic tests for non-commercial purposes. Rivers’ bill did not get out of committee. Dr. Leonard agrees with its underlying premise. “Diagnostic testing using DNA sequences is much more analogous to a medical practice than to manufacture,” she says. A bill introduced in 2007 would have prohibited patents on “a nucleotide sequence, or its functions and correlations, or the naturally occurring products it specifies.” It, too, was stillborn.

At the workshop, Dr. Ledbetter of Emory said gene patenting is not the main culprit in limiting availability of genetic tests. Rather, it is exclusive licenses issued by academic institutions where genes are discovered. (About two-thirds of patented genes involved in exclusive licensing for DNA diagnostics were discovered at universities.) He urged universities and scientists to seek nonexclusive licenses, as recommended by the National Institutes of Health and the Association of University Technology Managers. “A practical goal is to have at least two labs offering every genetic test,” he told CAP TODAY.

Dr. Bale of GeneDx would like to see more scientists get involved in licensing. “Scientists do science,” she says. “They rarely get involved after it turns into business. This is unfortunate. If more scientists took an interest in what happens to their discoveries, it is highly likely that exclusive licenses wouldn’t be issued.”

However, things might not be that easy. “We did a survey of our members last fall asking about exclusive licenses,” says Dr. Sauer of BIO. “They expressed a great desire to get exclusive licenses.” Many said that when an exclusive license is not an option, that is a deal breaker.

Recommendations for legislation were made by the SACGHS Task Force on Gene Patents and Licensing Practices, chaired by James P. Evans, MD, PhD, Bryson professor of genetics and medicine, University of North Carolina. The task force found that patents and exclusive licenses compromise test quality and availability of second opinions. Three other major findings from its four-year effort were as follows:

  • In the diagnostic area, gene patents are doing little work in promoting progress. Rather, they are limiting choice and access to testing. “Over and over we saw that numerous labs would offer a genetic test,” Dr. Evans says. “It was only after invoking intellectual property [IP] rights that the market was cleared and choices narrowed.” Evidence for this statement came from eight case studies of specific genetic conditions, especially testing for hereditary hemochromatosis (in its early years), BRCA1 and 2, Canavan disease, spinocerebellar ataxia, and APOE for Alzheimer disease. In contrast, discoverers of genes for cystic fibrosis and Huntington disease patented the genes but chose to license them broadly. More than 50 U.S. laboratories offer testing for each of these conditions. “A lack of patentenabled exclusivity will do nothing but enhance the genetic testing landscape,” Dr. Evans concludes.

  • Current gene patent and licensing practices do cause harm. “We documented harm to people, such as those on Medicaid, from patent-enabled exclusivity,” Dr. Evans says. “When there are more providers, which is the rule where IP is not involved, there is a much better chance that those people will be covered.” Responding to criticism that the task force did not find “pervasive” harm, Dr. Evans says, “I think the bar for harm is set very low when one talks about patient health, especially where no benefit has been demonstrated.”

  • Gene patents threaten multiplex testing, which is a growing part of laboratory practice. “Fragmented ownership of the genome combined with the need to look at many genes to diagnose a disease or predisposition sets up a severe potential for blocking problems,” Dr. Evans says. “We documented that lab directors right now are not reporting out results that are medically meaningful for fear of infringement.”

    Fragmented ownership could obstruct an even more innovative and promising technology, whole-genome sequencing. “We foresaw major problems on the horizon with whole-genome sequencing,” Dr. Evans says. “Not only will labs have to pay royalties for each gene, but in some cases simply knowing who owns a gene is difficult to ferret out. Gene patents will amplify costs of whole-genome sequencing tremendously.” So much for the promised $1,000 genome.
  • The task force recommended two main legis­lative changes: an exemption from infringement liability for those who use, make, or order diagnostic genetic tests in the care of patients; and an exemption from infringement liability for research on a patented gene.

    Dr. Evans would also like to see a move beyond voluntary compliance with recommendations for nonexclusive licensing. One option would be to make adherence to such principles a component in NIH grant funding. That would affect most, but not all, gene discoveries.

    Dr. Evans emphasizes that these recommendations are narrowly focused. “They would affect only the diagnostic arena and would leave patent incentives for the development of therapies intact,” he says. “One can make a strong argument that the development of therapeutics is so expensive that patent incentives are necessary, or at least desirable.” Whether victory in the AMP/ACLU lawsuit would have a narrowly focused impact is not clear. “If the plaintiffs won a broad victory, all gene patents would be invalidated, regardless of what use they were put to,” Dr. Evans says, and adds: “I’m no legal scholar, but I’m pretty sure that the scope of the ultimate ruling could vary from narrow to broad. I suspect it is misleading to speak in terms of ‘winning’ or ‘losing.’”

    As for whether changes will decrease investment, Dr. Evans is skeptical. “Where nonexclusivity prevails we still see robust investment,” he says. “Our changes will indeed adversely impact those companies whose business model relies simply on patent-enabled exclusivity rather than on enhanced innovation and quality. But the patent system was not created to simply support business models.”

    The eight case studies that the task force relied on were prepared by a team of researchers led by Robert Cook-Deegan, MD, director of the Center for Genome Ethics, Law and Policy, Duke University. “There is a lot of discussion in this area but most of it is not grounded in empiricism and is frankly pretty ideological,” says Dr. Cook-Deegan, who trained as a molecular biologist. Having real-world case studies provided a factual basis for the task force’s deliberations.

    One revealing study concerned the availability of testing for genes underlying two inherited colo­rectal cancer (CRC) syndromes: hereditary nonpoly­posis colon cancer (HNPCC, Lynch syndrome) and familial adenomatous polyposis (FAP). Between five and 10 laboratories offer testing for the various genes involved in each syndrome. “The upshot is that there are patents on many of those FAP and HNPCC genes,” Dr. Cook-Deegan says. “At the same time, there are many labs testing for them and listed on, which is highly public. So we presume they have nonexclusive licenses.” (Licensing agreements are confidential.) Dr. Cook-Deegan concludes that the difference between wide availability of testing with colorectal cancer genes, with Myriad as one option, and testing for BRCA genes, for which Myriad is the only option, is not the presence of patents, but how the patents are licensed.

    Dr. Cook-Deegan agrees that gene patents could threaten whole-genome sequencing. “The legacy of claims that would preclude anyone from deriving anyone’s DNA sequence for diagnostic purposes could hang over this growing technology,” he says. But he doesn’t see that happening. “I think the business dynamics won’t work out if those patents are strictly enforced,” he says. “And I don’t think courts will tolerate that.” He foresees that genome-sequencing companies would pay patent holders a small fee each time they run across a particular mutation. Navigenics has already proposed this arrangement (

    Like Dr. Evans, Dr. Cook-Deegan acknowledges that gene patents can be beneficial in the development of gene-based biopharmaceuticals. Asked whether success of the AMP/ACLU suit could eliminate this benefit, he replies that it could, but he doubts the court will go that far. “I could be wrong, but I predict that any judge handling a case will find a way to distinguish use of a gene for extracting information, such as for diagnosis or prognosis, from DNA used as a therapy itself or to make something, such as by gene transfer,” he says.

    “What struck us from the case studies,” Dr. Cook-Deegan says, “is that the genetic diagnostic market is radically different from therapeutics.” In most of its case studies, the team saw that while an exclusive license to a particular lab does seem to induce investment by that lab in getting a product to market, in all those cases there were already tests for the condition on the market. “What the patent did was to clear the market of alternative tests. When that is the situation it simply cannot be true that patents are necessary to get the diagnostic product to market. Those who say that we won’t have these tests if we don’t have patents and exclusive licenses seem to be wrong in all the real-world examples we examined.”

    Where exclusive licenses already exist, Dr. Cook-Deegan would like to see explicit provision for second opinions. “Companies like Myriad and Athena could get themselves out of the crosshairs by merely announcing that they permit verification testing as a matter of policy,” he says. “It would only come up in cases where they have already gotten their money from a patient, so it would not affect their market.”

    Dr. Sauer of BIO, commenting on the task force’s report from the industry viewpoint, says: “In our criticism of the report and its preliminary conclusions, maybe we could have been more explicit in complimenting the task force on a range of excellent work and detailed and fair case studies. We will all draw on that work for years to come.” That said, he adds, “Two aspects of the task force’s report bother us.” First, “I don’t think its findings unequivocally support its conclusions. Some evidence in the case studies seems to us to support gene patents even for diagnostic testing. There are instances where diagnostic tests sprung up spontaneously once a gene sequence with medical implications was published. What patents are useful for is higher investment in that technology.” Dr. Sauer cites Myriad’s claim that patent protection enabled its $200 million investment in educating physicians and patients and in genetic counseling.

    A second problem: “We don’t think the legislative changes recommended would be productive in the current climate,” Dr. Sauer says. Major patent reform bills were pending in the last three Congresses, he notes. “The debate about patent reform has become highly charged,” he says. “It is already difficult without adding the issue of genes.”

    Dr. Evans is skeptical about Myriad’s education campaign as a justification for gene patents. “As a practicing physician I have a pretty dim view of ‘educational efforts’ by companies,” he says. “With the robust nature of information exchange these days, I can sit in my office and review the entire world’s literature on any subject. So the desirability or even need for those with a huge vested interest in promoting their own product to be the ones to tell me about it eludes me.”

    As for genetic counseling, one might note that commercial labs like Quest and LabCorp have invested in extensive genetic counseling services for cystic fibrosis testing even without exclusive licenses.

    Congress, courts, and public policy forums—all are in play in the debate over gene patents. Dr. Sauer doesn’t hope for much from the court. “I think the outcome of the lawsuit is going to be unsatisfying to everyone involved,” he says. “It addresses narrow patent law issues. The court’s decision will say whether genes are patentable, but not whether they should be patentable. Congress may be a better area for settling this issue.”

    Dr. Caplan hopes for something positive from the lawsuit, regardless of its outcome. In his online commentary he wrote: “Perhaps the ACLU lawsuit, as unlikely to succeed as I think it is, will achieve its goal by reminding the emerging world of genetic medicine that, as Benjamin Franklin understood when he refused to patent his lightning rod, service to humanity is part of the duty of those who invent and innovate.”

    William Check is a medical writer in Wilmette, Ill.

    Related Links Related Links