YouTube Facebook LinkedIn Google+ Twitter Xingrss  

Does Whole Genome Sequencing Circumvent Gene Patents?


By W. Nicholson Price II 

December 10, 2012 | Guest Commentary | What happens when, during the course of whole-genome sequencing (WGS) a patient or research subject, an investigator sequences and analyzes a disease gene that has been patented? The U.S. Supreme Court will shed some light on this question next year when it issues its ruling in the long-running Myriad Genetics saga. 

Last month, the Supreme Court voted to hear the case of Association for Molecular Pathology v. Myriad Genetics to consider the question whether human genes are patentable. The plaintiffs—doctors, patients, researchers, and the American Civil Liberties Union—have challenged Myriad’s patents on the breast cancer genes BRCA1 and BRCA2, which cover, among other things, isolated DNA molecules with the sequences of those genes. A federal district court in New York ruled that the patent claims on isolated DNA molecules were invalid, but that ruling was reversed on appeal by the Federal Circuit in D.C. The Supreme Court decided to review the Federal Circuit’s decision and will likely rule on whether isolated human gene sequences are patentable next summer. 

This case has profound implications for biotechnology, and diagnostics, as well as the emerging field of personalized medicine. Among the fascinating issues that will likely be addressed is whether WGS—an essential foundation for truly personalized medicine—violates human gene patents.   

As WGS involves determining the sequence of an individual’s entire genome, there is concern in many quarters that WGS could violate essentially every patent covering an isolated human DNA sequence—of which there are thousands. Indeed, this concern has been raised by scholars, policy analysts and lawyers, including before the Federal Circuit and in the arguments over whether the Supreme Court should hear the case.  

However, a closer look at the technology suggests that rather than violating thousands of gene patents, WGS methods violate few, if any, existing gene patents. Whatever the Supreme Court decides next summer, the widespread adoption of clinical WGS is not particularly threatened by gene patents. 

Patents and WGS 

Gene patents, it is worth emphasizing, do not mean that the patentee owns the gene, in all forms and for all uses. Instead, a patentee has only the ability to prevent others from making or using what is specifically claimed in a patent; the patent claims determine the precise boundaries of the protected invention. While it is impossible to analyze comprehensively all the claims in all gene patents, an analysis by law professor Christopher Holman of a set of representative gene patents (including those at stake in the Myriad case) suggests that most claims in gene patents will not be infringed by WGS.   

The claims in gene patents generally fall into two categories: (1) composition of matter claims, which claim physical isolated DNA molecules; and (2) method claims, which claim methods of comparing the sequence of an individual with known reference sequences, and sometimes using that information to draw medical conclusions.   

Method claims are not especially problematic for WGS. Myriad’s method claims were held invalid by the district court; the Federal Circuit affirmed (relying on Mayo Collaborative Services v. Prometheus Laboratories, Inc., a March 2012 Supreme Court case which invalidated simple diagnostic correlation patents). Diagnostic methods patents thus have a somewhat uncertain future; it is unclear what genetic methods patents require to be valid. More importantly for the development of personalized medicine, diagnostic patents lack the same sort of potential hold-up problems as patents on isolated DNA. If any individual valid diagnostic patent covers a particular gene’s correlation with a disease and the patent is too hard or expensive for the person performing WGS to license, that result could just be left out of a WGS analysis without blocking the sequencing and the rest of the analysis.   

Composition claims, on the other hand, are the real source of worries for WGS and are squarely at issue in the Myriad case. Two characteristics of composition of matter claims are relevant: they claim physical DNA molecules which are (1) “isolated” and (2) generally quite long.  Because WGS typically does not make or use long isolated DNA molecules, it is unlikely to infringe composition claims. 

Composition claims in gene patents almost always claim “isolated” or “purified” DNA molecules with the specified gene sequence (or a set of related sequences). The claims are limited to isolated molecules because otherwise they would cover the gene as found in nature, and would thus be invalid as including unpatentable subject matter. The term “isolated” is ambiguous, and no court has laid out a generally applicable definition. (Indeed, since each patent can define its own term, the exact contours of “isolated” cannot be precisely determined for the whole class of gene patents.) However, “isolated” must have a relatively narrow meaning to avoid patent claims which are too broad and thus invalid. At the least, to avoid claiming genes as found in nature, “isolated” must mean that the claimed DNA is largely separated from other cellular components and flanking DNA sequences. 

The second relevant characteristic relates to the size of the claimed molecules. Most composition claims are for whole genes, or at least portions of genes long enough to encode functional proteins; in other words, composition claims generally cover lengthy DNA molecules of thousands of bases. (A handful of other claims cover all molecules containing a very short specified sequence—generally 15 or so bases—but these claims are almost certainly invalid as too broad and as anticipated by early public disclosures of matching DNA molecules.) 

In Violation 

To violate gene patents, then, WGS would need to make or use long, isolated DNA sequences. But most WGS techniques do not rely on such DNA molecules. Shotgun sequencing generally reads very short DNA molecules, frequently in the tens to hundreds of bases; while these molecules are isolated, they are usually too short to include the full sequence of most claimed DNA sequences. And next-generation WGS methods avoid even that possibility. Techniques that do not rely on PCR amplification almost certainly do not create “isolated” DNA molecules, and neither do techniques like the Pacific Biosciences platform or nanopore sequencing, which in theory sequences very long molecules of many kilobases—much longer than the approximately gene-length sequences claimed by most gene patents.  

Note that isolating the informational sequence of a particular gene—after the whole genome has been sequenced—does not violate composition claims because those claims cover physical molecules, and WGS never makes or uses the physical isolated DNA molecule containing that sequence. Thus, while there may be some residual uncertainty in this analysis from the sheer number of gene patents and their variation—and, as Robert Cook-Deegan has argued in Science, uncertainty itself can prevent innovative technologies from being adopted—it seems that WGS, especially next-generation WGS, likely infringes very few if any patents on isolated human gene sequences. 

(The same analysis applies to whole-exome sequencing (WES), even though initially the technique appears more likely to violate composition claims. However, as in WGS, the genome is broken into very short fragments prior to WES. The method thus never makes or uses the physical isolated gene-length DNA fragments claimed in gene patent composition claims, and likely infringes as few patents as WGS.) 

In addition to the legal analysis, there are some practical considerations that make it unlikely that gene patents will block widespread WGS. Most significantly, WGS has been going on for several years, and there have been no reports of specific gene patent holders trying to extract licenses or block WGS. Indeed, given that gene patents cover only very small portions of the genome, any successful infringement lawsuit would likely result only in small required royalty payments, rather than an injunction blocking the sequencing, making such a lawsuit not worth the time and cost of bringing it. Finally, of course, patents are granted for a limited time, many gene patents have expired, and it is much harder to get new gene patents. Gene patents, while still important, are a dying breed. 

Overall, several important issues may turn on the outcome of the closely-followed Myriad case.  But the future of whole-genome sequencing is not one of them. 

Editor’s note: This article is based on a more detailed analysis by the author published in 2012 in the Cardozo Law Review. 

Nicholson Price is an Academic Fellow at the Petrie-Flom Center for Health Law Policy, Biotechnology and Bioethics at Harvard Law School.  He can be reached at wprice@law.harvard.edu. 

Click here to login and leave a comment.  

2 Comments

  • Avatar

    The real question the Holman paper fails to address is how many of the gene patents with sliding oligo claim language like Myriad claim 5 have subject matter which would support divisionals and or a continuation in part (CIP). The most damaging position for a clinical sequencing lab would be claim language in the 100bp similarity range as its the readlength common to most clinical sequencing instruments and its specific enough to not be shot down for being too broad. I suspect most gene patents have this supporting subject matter and the cost to draft new claim language in a CIP is small compared to the money the assignees have already sunk into the estate.

    Robert Cook-Deegan hit the nail on the head. Its the uncertainty, rational or not, which is the poison. VC funding is down significantly in the life sciences this year despite Morgan Stanley being very bullish on the clinical sequencing space.
    http://www.genomeweb.com/vc-funding-life-sciences-plummets-q2-report-finds

  • Avatar

    As a person involved in clinical sequencing, I would caution this individuals advice.
    Currently claim 5 of the Myriad patent is in force (The 15bp of similarity claim). It will require substantial resources for a clinical lab to challenge this validity in court. I agree its too broad as BRCA1 has an Alu in exon 4 and this language lays claim to much more than the patent anticipates but it is nontheless in force today in the jurisdiction of the patent.

    2)You cant have it both ways. Short sequences dont violate because they dont cover the whole sequence ID (Not true in the case of Myriad claim 5) and long sequences dont violate because they dont explain the exact sequence ID (not true in Myriad claim 5). Which is it? Today, I believe both violate as Judge Louries discussion on DNA isolation was anchored in the Parke-Davis trials purification of adrenaline where NO BONDs had to be broken. It just had to be isolated from contaminating molecules like proteins and histones. Further modification of the DNA to add primer sites (as is done in PacBio) would put one even closer to the DNA isolation claims of Myriad. Likewise PacBio today not only recombines the molecules to put primer sites on them, it also amplifies to target its sequencer for clinical samples as it does not have the economies to perform WGS or WES.

    In this regard it is incorrect to assert no one has utilized gene patents against next gen sequencing clinical facilities. Athena Diagnostics vs Ambry genetics is currently ongoing and I believe the SACGHS report has many other examples of Canavan disease, LongQT, and SCA being enforced against clinical labs. Most clinical labs have opted to drop the gene from their panels than fight the legal costs of the battle. This has a negative impact on patients seeking a second opinion.

    The real question the Holman paper fails to address is how many of the gene patents with sliding oligo claim language like Myriad claim 5 have subject matter which would support divisio

Add Comment

Text Only 2000 character limit

Page 1 of 1

For reprints and/or copyright permission, please contact  Jay Mulhern, (781) 972-1359, jmulhern@healthtech.com.