Illumina, Inc. v. Ariosa Diagnostics, Inc. (Fed. Cir. 2020)

Federal Circuit Hands Down Modified Opinion in Illumina, Inc. v. Ariosa Diagnostics, Inc.

By Kevin E. Noonan --

Earlier this year, the Federal Circuit (somewhat surprisingly) found claims of two Sequenom patents directed to methods for detecting fetal DNA in maternal blood to satisfy the subject matter eligibility requirements of Section 101 (see "Illumina, Inc. v. Ariosa Diagnostics, Inc. (Fed. Cir. 2020)").  The surprise arose in part due to the Federal Circuit's track record of finding all diagnostic method claims to be ineligible as being directed to a natural law without "something more" to overcome the patentability preclusion created by the Supreme Court in Mayo Collaborative Services v. Prometheus Laboratories, Inc. (see, e.g., Judge Moore's dissent in Athena Diagnostics, Inc. v. Mayo Collaborative Servs., LLC, 927 F.3d 1333, 1352 (Fed. Cir. 2019) ( "Since Mayo, we have held every single diagnostic claim in every case before us ineligible.").  Another reason, of course, is that the Federal Circuit affirmed a finding of patent ineligibility in Ariosa Diagnostics, Inc. v. Sequenom, Inc.  Nevertheless, a divided panel found a patent eligibility-creating distinction in the claims asserted in this most recent case (over a dissent by the author of the Ariosa v. Sequenom decision, Judge Reyna) and today the Court issued a revised opinion in the face of Ariosa's petition for rehearing.

To recap, the case arose over U.S. Patent No. 9,580,751 (the '751 patent) and U.S. Patent No. 9,738,931 (the '931 patent), directed to the solution of an unexpected difficulty in detecting cell-free fetal DNA (cffDNA):

[T]he major proportion (generally >90%) of the extracellular DNA in the maternal circulation is derived from the mother.  This vast bulk of maternal circulatory extracellular DNA renders it difficult, if not impossible, to determine fetal genetic alternations [sic] . . . from the small amount of circulatory extracellular fetal DNA.

The inventors found that cffDNA was significantly smaller (300-500 bp) than the "interfering" maternal DNA, and thus using admittedly conventional techniques of size separation the cffDNA could be isolated and rendered detectable.

Illumina and Sequenom asserted claims 1, 2, 4, 5, 9, and 10 of the '751 patent and claims 1, 2, and 10–14 the '931 patent against Ariosa and Roche Diagnostics; the Court considered claim 1 from each patent to be representative:

'751 patent:

1. A method for preparing a deoxyribonucleic acid (DNA) fraction from a pregnant human female useful for analyzing a genetic locus involved in a fetal chromosomal aberration, comprising:
   (a) extracting DNA from a substantially cell-free sample of blood plasma or blood serum of a pregnant human female to obtain extracellular circulatory fetal and maternal DNA fragments;
       (b) producing a fraction of the DNA extracted in (a) by:
           (i) size discrimination of
           (ii) selectively removing the DNA fragments greater than approximately 500 base pairs,
           wherein the DNA fraction after (b) comprises a plurality of genetic loci of the extracellular circulatory fetal and maternal DNA; and
       (c) analyzing a genetic locus in the fraction of DNA produced in (b).

'931 patent:

1. A method, comprising:
   (a) extracting DNA comprising maternal and fetal DNA fragments from a substantially cell-free sample of blood plasma or blood serum of a pregnant human female;
       (b) producing a fraction of the DNA extracted in (a) by:
           (i) size discrimination of extracellular circulatory fetal and maternal DNA fragments, and
           (ii) selectively removing the DNA fragments greater than approximately 300 base pairs,
       wherein the DNA fraction after (b) comprises extracellular circulatory fetal and maternal DNA fragments of approximately 300 base pairs and less and a plurality of genetic loci of the extracellular circulatory fetal and maternal DNA fragments; and
       (c) analyzing DNA fragments in the fraction of DNA produced in (b).

The District Court on summary judgment held these claims and all asserted claims to be ineligible under 35 U.S.C. § 101 and the Supreme Court's Alice/Mayo test (see "Illumina, Inc. v. Ariosa Diagnostics, Inc. (N.D. Cal. 2018)"), and Illumina appealed.

In the original opinion, the Federal Circuit reversed, in an opinion by Judge Lourie joined by Judge Moore; Judge Reyna dissenting (as he does in this revised opinion).  Judge Lourie justified the different outcome here by stating "[t]his is not a diagnostic case.  And it is not a method of treatment case.  It is a method of preparation case."  But the majority based its decision on this distinction:  "[h]ere, it is undisputed that the inventors of the '751 and '931 patents discovered a natural phenomenon.  But at step one of the Alice/Mayo test, 'it is not enough to merely identify a patent-ineligible concept underlying the claim; we must determine whether that patent-ineligible concept is what the claim is 'directed to," citing Rapid Litig. Mgmt. Ltd. v. CellzDirect, Inc.  It is undisputed that cell-free fetal DNA (cffDNA) exists in maternal blood and its presence is a natural phenomenon, and moreover that the size differentials and distributions that form the predicate basis for the claims at issue here are also naturally occurring.  But the panel majority held that the claims were not merely "directed to" that natural phenomenon, but rather were directed to a method that exploits the natural phenomenon that renders the claims patent-eligible.  Sounding a theme he first enunciated in Vanda Pharmaceuticals Inc. v. West-Ward Pharmaceuticals, Judge Lourie recognizes the distinction that the claims at issue recited "specific process steps—size discriminating and selectively removing DNA fragments that are above a specified size threshold" that increased the relative amount of cffDNA in the processed sample compared to maternal DNA and thus "change[d] the composition of the mixture" and produced a DNA-containing fraction that was different from what naturally occurs in maternal blood.

Judge Lourie's latest opinion reiterates most of these themes.  The basis of the decision remains that "this is not a diagnostic case.  And it is not a method of treatment case.  It is a method of preparation case."  As in the original opinion, the panel majority are somewhat critical of appellants' inability (in the majority's view) to "clearly identify the natural phenomenon that forms the basis of its challenge" but finds that the parties' differences on this point to be relatively meaningless distinctions and adopts Illumina's definition that the natural phenomenon underlying the claimed invention is that "cell-free fetal DNA tends to be shorter than cell-free maternal DNA in a mother's bloodstream."  But the majority arrives at the same conclusion as before:  although the claims are based on this natural phenomenon they "are not directed to that natural phenomenon but rather to a patent-eligible method that utilizes it" (emphasis in opinion).

And the basis of this conclusion relies at least in part on the specific method steps recited by the claims, which might constitute the ineluctable "something more" mandated by the Supreme Court's jurisprudence that looks enough like novelty to disturb any patent practitioner (or more likely scholar) concerned with doctrinal consistency.  But this insistence also cabins the scope of the claims to the specifically recited steps, which goes a long way towards avoiding the overbroad scope that seems to have been at least part of the Supreme Court's concern regarding preemption motivating the Court's attempts to restrict patent eligibility of certain types of claims.  Also important for the panel majority is that the choice of the size distinctions recited in the claims are not themselves natural phenomenon but are "human-engineered parameters that optimize the amount of maternal DNA that is removed from the mixture and the amount of fetal DNA that remains in the mixture in order to create an improved end product that is more useful for genetic testing than the original natural extracted blood sample."  And the result of the claimed methods are a mixture of DNA fragments having a change in their composition, being enriched in fetal-derived DNA fragments shorter than 500 (the '751 patent) or 300 (the '931 patent) basepairs.

The opinion, like its earlier version, distinguishes the claims at issue here with the claims in Supreme Court and Federal Circuit precedent found to be patent-ineligible, most notably that these claims are not directed to isolated DNA itself (Association for Molecular Pathology v. Myriad Genetics, Inc., 569 U.S. 576 (2013)), or merely detecting its presence in maternal blood (Ariosa), or even determining "whether a cell-free DNA fragment in a previously-prepared sample is fetal or maternal based on the natural size distribution of cell-free DNA fragments" but rather methods for preparing a mixture enriched in cffDNA based on "conventional separation technologies can be used in unconventional ways," specifically the claimed thresholds of 500 bp and 300 bp for separating fetal from material DNA ("Roche[] has presented no evidence that thresholds of 500 base pairs and 300 base pairs were conventional for separating different types of cell-free DNA fragments.").

An admittedly cursory comparison between the majority opinion issued in March and the one issued today finds precious little difference in the reasoning advanced by the Court.  In contrast, Judge Reyna's dissent appears to refine some of the Judge's arguments but at bottom reiterates his opinion that these claims are directed to a patent-ineligible natural phenomenon.

The Court is properly not transparent regarding the internal discussions and arguments asserted by those Judges who share the majority's views and those who share Judge Reyna's views.  But in light of the great similarities of these opinions it seems apparent that the Court was most comfortable not deciding to hear the case as a result of the revisions occasioned in the opinion handed down today.  Whether there are additional motivations (including putting the panel's disparate views in better condition for Supreme Court review) may become apparent if the Court deigns to revisit what it has wrought in the years since deciding the proper metes and bounds of Section 101 became a priority for the Court.

Illumina, Inc. v. Ariosa Diagnostics, Inc. (Fed. Cir. 2020)
Panel: Circuit Judges Lourie, Moore, and Reyna
Opinion by Circuit Judge Lourie; dissenting opinion by Circuit Judge Reyna

Comments

[skeptical]

That was a ton of words without actually delineating what the revision to the decision actually was.

I admit that somewhere in there the point of the revision may have been included, but the search party looking for it now needs another search party to find it.

[Paul Cole]

As Francisco said to Barnardo in Act 1 of Hamlet: "For this relief much thanks."

[Paul F Morgan]

Thanks for noting the amended decision statement that: ("Roche[] has presented no evidence that thresholds of 500 base pairs and 300 base pairs were conventional for separating different types of cell-free DNA fragments."). I.e., the method specifically claimed is apparently novel.

[Kevin E Noonan]

Skeptical: I could find the following differences between the two opinions:

March: The claims in this case are directed to methods for pre- paring a fraction of cell-free DNA that is enriched in fetal DNA. The methods include specific process steps—size discriminating and selectively removing DNA fragments that are above a specified size threshold—to increase the relative amount of fetal DNA as compared to maternal DNA in the sample. ’751 patent col. 7 ll. 63–67. Those process steps change the composition of the mixture, resulting in a DNA fraction that is different from the naturally-occurring fraction in the mother’s blood. Thus, the process achieves more than simply observing that fetal DNA is shorter than maternal DNA or detecting the presence of that phenome- non.

August: The claims in this case are directed to methods for pre- paring a fraction of cell-free DNA that is enriched in fetal DNA. The methods include specific process steps—size discriminating and selectively removing DNA fragments that are above a specified size threshold—to increase the relative amount of fetal DNA as compared to maternal DNA in the sample. ’751 patent col. 7 ll. 63–67. The size thresholds in the claims—500 base pairs in the ’751 patent and 300 base pairs in the ’931 patent—are not dictated by any natural phenomenon, particularly because the size distributions of fetal and maternal cell-free DNA overlap each other (i.e., there are maternal DNA fragments shorter than 300 base pairs). The claimed size thresholds are human-engineered parameters that optimize the amount of maternal DNA that is removed from the mixture and the amount of fetal DNA that remains in the mixture in order to create an improved end product that is more useful for genetic testing than the original natural extracted blood sample.Moreover, the claimed methods achieve more than simply observing that fetal DNA is shorter than maternal DNA or detecting the presence of that phenomenon. The claims include physical process steps that change the composition of the mixture, resulting in a DNA fraction that is different from the naturally occurring fraction in the mother’s blood.

March: Here, in contrast, the claims are directed to more than just the correlation between a DNA fragment’s size and its tendency to be either fetal or maternal. And the claims do not merely cover a method for detecting whether a cell-free DNA fragment is fetal or maternal based on its size. Rather the claimed method removes some maternal DNA from the mother’s blood to prepare a fraction of cell-free DNA that is enriched in fetal DNA. Thus, the claims in this case are different from the claims that we held invalid in Ariosa.

August: Here, in contrast, the claims are directed to more than just the correlation between a DNA fragment’s size and its tendency to be either fetal or maternal, a correlation which is not even mentioned in the claims. The claims do not cover a method for detecting whether a cell-free DNA fragment in a previously-prepared sample is fetal or ma- ternal based on the natural size distribution of cell-free DNA fragments; rather, the claimed methods exploit that natural size distribution during the sample preparation steps to remove some maternal DNA from the mother’s blood. Even the “analyzing” step of the claims is not di- rected to analyzing the discovered natural phenomenon, but to analyzing something else entirely, namely, “fetal chromosomal aberrations.” See ’751 patent col. 7 ll. 55–56, col. 8 ll. 56–57, col. 9 ll. 5–8; ’931 patent col. 9 ll. 17–24. Thus, the claims in this case are different from the claims that we held invalid in Ariosa.

March: The Court stated:
It is important to note what is not implicated by this decision. First, there are no method claims be- fore this Court. Had Myriad created an innovative method of manipulating genes while searching for the BRCA1 and BRCA2 genes, it could possibly have sought a method patent. But the processes used by Myriad to isolate DNA . . . are not at issue in this case.

August: The Court stated:
It is important to note what is not implicated by this decision. First, there are no method claims before this Court. Had Myriad created an innovative method of manipulating genes while searching for the BRCA1 and BRCA2 genes, it could possibly have sought a method patent. But the processes used by Myriad to isolate DNA were well understood by geneticists at the time of Myriad’s patents, were well understood, widely used, and fairly uniform insofar as any scientist engaged in the search for a gene would likely have utilized a similar approach, and are not at issue in this case.

August: Here, we encounter the opposite situation, i.e., the claims do not cover separated cell-free fetal DNA itself but rather a process for selective removal of non-fetal DNA to enrich a mixture in fetal DNA. That process includes size parameters that the inventors engineered to balance the practicalities of the specific problem that they were facing, namely, removing enough cell-free maternal DNA to enrich the mixture while leaving enough cell-free fetal DNA to allow for testing. Thus, the Supreme Court’s decision in Myriad is not on point in this case where the inventors claimed to have conceived and reduced to practice, not the separated DNA, but a method that uses unconventional size pa- rameters to perform the separation.

Except for the inclusion of the last paragraph above, there is little real difference between the two opinions. And while adding to the distinctions the court made between this case and Myriad (and its earlier decision in Ariosa v. Sequenom, I don't think these differences change the scope of the opinion.

If you can add anything please do.

[skeptical]

Thank you Dr. Noonan - any idea on why a second recent case has been 'amended?'

[Kevin E Noonan]

That's next

[skeptical]

Patience is not my strong suit, Dr. Noonan.

When is 'next?'

[skeptical]

It has arrived - thank you!