By Kevin E. Noonan --
The Federal Circuit returned to its consideration of the outcome in the District Court of The Medicines Company's ANDA litigation against Mylan and Bioniche Pharma over a proposed generic version of Medicines' bivalirudin drug (sold as Angiomax®), an anticoagulant used in heart surgery. In doing so, the Court focused on claim construction, basing its decision to reverse a finding of infringement under 35 U.S.C. § 271(e)(2) on differences with the District Court's claim construction.
Claim 1 of the '727 patent:
1. Pharmaceutical batches of a drug product comprising bivalirudin (SEQ ID NO: 1: [Phe Pro Arg Pro Gly Gly Gly Gly Asn Gly Asp Phe Glu Glu Ile Pro Glu Glu Tyr Leu]) and a pharmaceutically acceptable carrier for use as an anticoagulant in a subject in need thereof, wherein the batches have a pH adjusted by a base, said pH is about 5-6 when reconstituted in an aqueous solution for injection, and wherein the batches have a maximum impurity level of Asp9-bivalirudin that does not exceed about 0.6% as measured by HPLC.
Claim 1 of the '343 patent:
1. Pharmaceutical batches of a drug product comprising bivalirudin ... prepared by a compounding process comprising:
(i) dissolving bivalirudin in a solvent to form a first solution;
(ii) efficiently mixing a pH-adjusting solution with the first solution to form a second solution, wherein the pH-adjusting solution comprises a pH-adjusting solution solvent; and
(iii) removing the solvent and pH-adjusting solution solvent from the second solution;
wherein the batches have a pH adjusted by a base, said pH is about 5-6 when reconstituted in an aqueous solution for injection, and wherein the batches have a maximum impurity level of Asp9-bivalirudin that does not exceed about 0.6% as measured by HPLC.
The patents-in-suit arose from Medicines' solution to production problems with the drug, which led to the production of a degradation product, Asp9 bivalirudin (caused by deamination of Asn9); in high enough concentrations (>1.5%) Asp9 bivalirudin contaminants were unacceptable (per FDA). Production of bivalirudin batches contaminated with high levels of Asp9 bivalirudin caused Medicines to shut down commercial production for more than a year. The company's attempts to produce formulations of the drug for commercial sale were unsuccessful, and led Medicines to develop certain ways of adjusting the pH that reduced the impurity levels to less than 0.6% and resulted in the '343 and '727 patents.
These patents were the subject of litigation between Medicines and Hospira, where coincidentally the same panel that rendered the instant decision determined that the claims were invalid under the on-sale bar of 35 U.S.C. § 102(b)(among other grounds) (see "The Medicines Company v. Hospira (Fed. Cir. 2015)"). The Federal Circuit thereafter vacated the panel opinion and reinstituted the appeal en banc, ordering the parties (and amici) to brief issues relating to the on-sale bar and its application to the relationship between Medicines and its contract manufacturer. The en banc court reversed the decision by the merits panel that the on-sale bar applied, in a unanimous court by Judge O'Malley (see "The Medicines Company v. Hospira, Inc. (Fed. Cir. 2016) (en banc)").
The Federal Circuit (which had dismissed Medicines' and Mylan's cross-appeals in view of the merit panel's decision in the Hospira case) then reinstated this appeal between Medicines and Mylan. The District Court had granted summary judgment of non-infringement to Mylan as to the claims of the '343 patent, based on Mylan's ANDA specifying methods for formulating that did not use the claimed "efficient mixing" limitation of that patent's claims. After a bench trial, the District Court had granted judgment in Medicines' favor, expressly because the '727 patent claims did not recite the "efficient mixing" limitation.
The Federal Circuit affirmed the District Court's summary judgment decision regarding the '343 patent claims, but reversed the judgment regarding the '727 patent claims, in an opinion by Judge Dyk joined by Judges Hughes and Wallach. The Court expressly based its decision on construction of the term "batches" in the '727 patent claims, which required the use of efficient mixing methods as disclosed in the specification shared in common by both patents. This specification defined "batches" as:
As used here, "batch" or "pharmaceutical batch" refers to material produced by a single execution of a compounding process of various embodiments of the present invention. "Batches" or "pharmaceutical batches" as defined herein may include a single batch, wherein the single batch is representative of all commercial batches (see generally, Manual of Policies and Procedures, Center for Drug Evaluation and Research, MAPP 5225.1, Guidance on the Packaging of Test Batches at 1), and wherein the levels of, for example, Asp9-bivalirudin, total impurities, and largest unknown impurity, and the reconstitution time represent levels for all potential batches made by said process. "Batches" may also include all batches prepared by a same compounding process.
The District Court construed the term "pharmaceutical batches" consistent with specification, to be either a single batch representative of all commercial batches "made by a compounding process" having consistent levels of all impurities (including Asp9-bivalirudin) or all batches made by the compounding process; the Federal Circuit opinion notes that the parties consented to this construction. With regard to the term "efficient mixing," the District Court based its construction on two examples in the common specification: one the "inefficient" mixing according to earlier Medicines' processes (Example 4), and the efficient process disclosed in the specifications (Example 5); the District Court further held that Medicines had disclaimed the old inefficient mixing process.
With regard to its review of claim construction, the opinion notes that claim construction here is reviewed as a matter of law because the District Court did not consider anything outside the intrinsic evidence, under the Supreme Court's Teva Pharmaceuticals v. Sandoz decision. The panel based its contrary decision on the District Court's failure to interpret the "batches" limitation to require efficient mixing for the claims of the '727 patent. The panel's rationale was that the "batches" limitation requires the claims to encompass only those having impurity levels less than 0.6%, and that this limitation cannot cover individual batches because it was undisputed that such batches were known in the prior art (i.e., batches Medicines made prior to the disruption that led to this invention). The opinion states: "[r]ather, properly construed, what the batches limitation requires is the use of a process that achieves batch consistency," expressly construing the "batches" claim limitation in view of the specification. "The batches limitation therefore requires a process that achieves consistency between batches produced from the 'same compounding process'—i.e., batch consistency," said the Court. Indeed, the panel viewed this consistency as being what distinguished these claims from the prior art (including Medicines' own prior methods and batches). Without tying the interpretation of the term "batches" to the disclosed more efficient methods, the opinion asserts that there would be no way of effectively deciding when and whether infringement would lie. The opinion foresees a need for "forward-looking assessments of whether future batches would infringe" (i.e., only after the batches were tested to determine the level of Asp9-bivalirudin levels). According to the opinion:
To illustrate, if a defendant using the same compounding process produced fifty batches each having an Asp9 level below 0.6 percent, each of those fifty batches would infringe. But the defendant would not know whether any of the batches infringed until all fifty batches had been produced because if even one of those batches was determined to have an Asp9 level higher than 0.6 percent, none of the batches would infringe.
In this regard, the opinion posits that such an interpretation would not provide the "reasonable certainty" required by Nautilus, Inc. v. Biosig Instruments, Inc., 134 S. Ct. 2120, 2129 (2014).
The opinion further found support for the Court's interpretation in the specification and prosecution history (particularly with regard to the claims patentability over inter alia the Medicines' own prior art).
Having determined that proper construction of the "batch" limitation required the use of a compounding process that achieves batch consistency, the opinion then turned to whether this compounding process must entail "efficient mixing." Evidence for this requirement the Court found in the specification, particularly disclosure that "pH-adjusting solution will be efficiently mixed," and that "[e]fficient mixing of the pH-adjusting solution . . . will minimize levels of Asp9-bivalirudin." Moreover, the Court states that the specification disclosed no other method capable of producing such batches with the necessary consistency (relating to the comparison between the methods disclosed in Examples 4 and 5). The opinion also asserted that the District Court's determination that Medicines disclaimed inefficient mixing was not inconsistent with its determination that what the claims required was efficient mixing.
The opinion also rejected Medicines' contention that this interpretation would make the claims of the '343 patent "superfluous" (more accurately, implicate double patenting) by noting there may be overlap between certain of the claims but that the more specific recitations in the '343 claims distinguished them from the '727 claims. The opinion also asserted that adopting Medicines' construction would potentially encompass the prior art, and thus impermissibly extending Medicines' "monopoly" beyond the invention disclosed in the common specification.
The Court then turned to what is meant by "efficient mixing." This term is explicitly defined in the common specification to mean "'mixing [that] is characterized by minimizing levels of Asp9-bivalirudin in the compounding solution,' i.e., below 0.6 percent Asp9-bivalirudin in the intermediate solution." The panel rejected this interpretation, insofar as "it does not accord with the linguistic formula used by the patentee to signal the designation of other defined terms—including 'batches.'" By this it appears the Court meant to distinguish between how this statement was phrased and how other terms were expressly defined, i.e., where "the defined term [is] in quotation marks, followed by the terms 'refers to' or 'as defined herein.'" The Court also rejected Medicines' assertion that its phrase was definitional on the basis that is was "a mere recitation of the results obtained" rather than constituting a definition of what "efficient mixing" is, according to the opinion. Using Medicines' definition "would expand the scope of 'efficient mixing' to cover any way of mixing that achieves a compounding solution having an Asp9 level of less than 0.6 percent" in violation of the written description requirement, said the panel, citing Ariad Pharms., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1352–53 (Fed. Cir. 2010) (en banc). The Court found no art-recognized definition of the term "efficient mixing," and the other portions of the specification Medicines' argued disclosed alternative methods were "vague and unhelpful" (and were not relied upon by the District Court or the parties). Finally, the opinion (relying with some irony on Medicines' criticism of the District Court for defining "efficient mixing" with regard to what it is not, i.e., prior art methods) notes that the skilled artisan would rely upon the methods recited in Example 5 to understand what efficient mixing is. In the Court's opinion, this Example is the only description of efficient mixing in the specification and not just an embodiment. According to the panel, this Example provides "a clear 'objective standard by which to measure the scope of the term.'"
The opinion concluded that, in order to infringe a batch must be compounded using a efficient mixing process according to Example 5 for claims of either the '727 and '343 patents.
There is some logic underlying this decision, but it can be understood if Medicines' views the Federal Circuit's logic as imposing a definition from the specification onto the claims, contrary to the claim construction rubric that limitations should not be incorporated from the specification into the claims. This aspect of the decision brings to mind District Court Judge Vaughn Walker's discussion of how vexatious this rule can be, as set forth his dissent in Medegen MMS, Inc. v. ICU Medical, Inc. (Fed. Cir. 2008):
The issues before the Court presented "a conundrum often faced by district courts when construing" claim terms, citing the "twin axioms" of claim construction that produced the conundrum: "[o]n the one hand, claims 'must be read in view of the specification'," citing Markman, while "[o]n the other hand, it is improper to read a limitation from the specification into the claims," citing Arlington Indus. Inc. v. Bridgeport Fittings, Inc. The problem with these axioms, according to Judge Walker, is that "the axioms themselves seldom provide an answer, but instead merely frame the question to be resolved."
"As a jurist more accustomed to working on the front lines of patent litigation than reviewing decisions from above, it is my experience that claim construction -- determining how one of ordinary skill in the art would understand the patent at the time of invention -- often requires making fact-like determinations not well suited to appellate review."
The Supreme Court's decision in Teva v. Sandoz reversed the Federal Circuit's practice of giving no deference to factual underpinnings of claim construction as set forth in Cybor Corp. v. FAS Technologies, Inc. The instant decision shows that while the problems attendant upon failure to give factual deference have been diminished they have not been overcome entirely.
The Medicines Company v. Mylan, Inc. (Fed. Cir. 2017)
Panel: Circuit Judges Dyk, Wallach, and Hughes
Opinion by Circuit Judge Dyk