Merck v. Gilead: $2.54 billion verdict

A federal jury in Delaware awarded Merck $2.54 billion in royalties over Gilead's Sovaldi and Harvoni hepatitis C drugs for infringing Merck's patent. The jury awarded Merck a royalty rate of 10% on total sales of $25.4 billion. The verdict can be found here: http://bit.ly/2gRtMqw

AMGEN V. HOSPIRA: “AN ISOLATED . . . ISOFORM” IS LIMITED TO ONE, AND DEPENDENT CLAIM TO MULTIPLE ISOFORMS IS INVALID

In Amgen v. Hospira[1], a patent infringement lawsuit (U.S. Patent No. 5,856,298) over the biosimilar version of Amgen’s Epogen (epoetin alfa), the parties disputed the meaning of the term "an isolated ... isoform" in claim 1:

1. An isolated biologically active erythropoietin isoform having a single isoelectric point and having a specific number of sialic acids per molecule, said number selected from the group consisting of 1-14, and said isoform being the product of the expression of an exogenous DNA sequence in a non-human eucaryotic host cell.

8. A composition consisting essentially of two or three erythropoietin isoforms according to claim 1. 

Amgen proposed that "an isolated ... isoform" allows for mixtures of at least one isoform.  Hospira proposed that the language allows for mixtures of only one isoform. 

The District Court (Richard Andrews) first discussed the case law regarding the meaning of the term ‘a’ or ‘an’ in a claim: “That 'a' or 'an' can mean 'one or more' is best described as a rule, rather than merely as a presumption or even a convention. The exceptions to this rule are extremely limited: a patentee must 'evince[] a clear intent' to limit 'a' or 'an' to 'one'. . .  An exception to the general rule that 'a' or 'an' means more than one only arises where the language of the claims themselves, the specification, or the prosecution history necessitate a departure from the rule."

Despite stating that the exceptions to this rule (limiting ‘a’ or ‘an’ to only one) are extremely limited, the District Court found that the exception applied “because the plain language evinces a clear intent to claim only one isoform. The claim language reads ‘an isolated ... isoform.’ Plaintiff's reading would render the word ‘isolated’ superfluous. Plaintiff's reading would equate the phrase ‘an isolated .. . isoform’ with ‘an isoform.’”  The District Court further emphasized that Claim 1 was amended during prosecution to read "[a]n isolated ... isoform."  The purpose of the amendment was "to further clarify that an erythropoietin isoform represents a homogeneous preparation."

After finding that Claim 1 requires only one isoform, the District Court found Claim 8 to be invalid since it contradicts Claim 1 's limitation that the isoform is "isolated." Claim 8 requires a mixture "consisting essentially of two or three" isoforms.  Claim 8 thus improperly narrows Claim 1.

The District Court did not discuss the doctrine of claim differentiation, and how the multiple isoforms of Claim 8 would imply that Claim 1 is not limited to a single isoform. 

The Opinion can be found here:

http://www.ded.uscourts.gov/sites/default/files/opinions/rga/2016/november/15-839.pdf

---

[1] Amgen v. Hospira, No. 15-cv-839 (D. Del., filed Sept. 18, 2015).

HERCEPTIN IMMUNOCONJUGATE IPR (KADCYLA BIOSIMILAR INTER PARTES REVIEW)

In view of Hospira’s <http://biopharmapatent.blogspot.com/2016/12/hospira-has-filed-petition-for-inter.html> and Mylan’s http://biopharmapatent.blogspot.com/2016/11/mylans-herceptin-ipr-biosimilar-inter.html recent IPR (Inter Partes Review) petitions of patents covering the drug HERCEPTIN^®, we look back at a previously decided IPR (IPR2014-00676) proceeding relating to a patent that covers immunoconjugates of the drug HERCEPTIN^®. 

On October 27, 2015, the PTAB (Patent Trial and Appeal Board) found the claims of U.S. Patent No. 8,337,856, relating to immunoconjugates of an anti-ErbB antibody to be patentable.  The ‘856 Patent covers immunoconjugates of the humanized anti-ErbB2 antibody known as HERCEPTIN^®, linked to a maytansinoid toxin.  The Patent covers the drug KADCYLA^® (ado-trastuzumab emtansine).

The Petitioner (Phigenix Inc.) argued that an ordinary artisan would have had reason to substitute the mouse monoclonal antibody in the immunoconjugate of the prior art (Chari 1992) with huMAB4D5-8 (HERCEPTIN^®) in particular because it was known that (1) humanized antibodies were preferred over mouse counterparts for clinical applications, (2) huMAB4D5-8 had been FDA approved for use to treat breast tumors in humans, and (3) clinical studies indicated that huMAB4D5-8 worked well in combination with microtubule-directed chemotherapy agents for the treatment of breast cancer.

Patent Owner (Genentech, Inc. and ImmunoGen, Inc.) argued that the prior art indicated that HERCEPTIN^® -maytansinoid immunoconjugates would have been expected to exhibit unacceptable levels of toxicity in normal human liver tissue in patients.  The Patent Owner pointed to Pai-Scherf 1999, which describes a Phase I clinical study of human patients receiving an immunoconjugate (erb-38) comprising a portion of the anti-HER2 monoclonal antibody e23 fused to a truncated form of Pseudomonas exotoxin A. Although the Pai-Scherf group “initiated the study in humans based on ‘excellent antitumor activity and acceptable animal toxicities,’” it nonetheless observed unacceptable hepatotoxicity in all patients in the treatment group.  Pai-Scherf 1999 indicates that, in a clinical study, human patients experienced “hepatic injury” when exposed to erb-38.  Pai-Scherf 1999 discloses that the “toxicity of erb-38 is most likely due to the presence of erbB2 on hepatocytes, not detected by immunohistochemical staining in earlier publications.”

In response, Petitioner contended that Pai-Scherf 1999 was not relevant because it described the use of a “fusion protein,” not an antibody-drug conjugate.  The PTAB disagreed, stating that although the reference disclosed clinical studies using a single-chain immunoconjugate comprising a portion of an antiHER2/erbB2 antibody and a truncated form of a toxin, Pai-Scherf 1999 discusses generally the “targeting of tumors with antibodies to erbB2 armed with radioisotopes or other toxic agents.” 

The PTAB held that the ordinary artisans would not have had a reasonable expectation that any immunoconjugate, much less the claimed Herceptin^®- maytansinoid immunoconjugate in particular, would be useful to treat solid tumors in humans. 

HERCEPTIN PROTEIN PURIFICATION IPR (BIOSIMILAR INTER PARTES REVIEW)

Hospira has filed a petition for an IPR (IPR2016-0183) (inter partes review) of Genentech’s U.S. Patent No. 7,807,799, which is directed to methods of purifying antibodies.  The claims of the '799 Patent cover any protein that has a CH2/CH3 region, including Genentech’s trastuzumab (Herceptin^®).  The Patent Trial and Appeal Board (PTAB) has not yet made a decision on the petition.  The ’799 Patent claims a method of purifying a protein which comprises a CH2/CH3 region by carrying out protein A affinity chromatography at a temperature in the range of about 10 °C to about 18 °C.  Hospira mainly relies on the following two references for teaching the limitations of the claims of the ‘799 Patent:

Claim Limitations	WO 95/22389
1. A method of purifying a protein which comprises a CH2/CH3 region	A method for the purification of an IgG antibody . . .
comprising subjecting a composition comprising said protein to protein A affinity chromatography	A method for the purification of an IgG antibody . . . comprising sequentially subjecting the medium to (a) Protein A affinity chromatography.
at a temperature in the range from about 10 °C to about 18 °C.	The process in its most preferred embodiment consists of three purification steps (Protein A affinity, cation exchange, and hydrophobic interaction chromatography) . . . All steps are carried out at room temperature (18 – 25 °C).
5. The method of claim 1 wherein the protein is an antibody.	A method for the purification of an IgG antibody . . .

Claim Limitations	Van Sommeren (22 Preparative Biochemistry 135 (1992))
1. A method of purifying a protein which comprises a CH2/CH3 region	The purification of immunoglobulins (IgG) . . .”
comprising subjecting a composition comprising said protein to protein A affinity chromatography	The purification of immunoglobulins (IgG), in particular mouse monoclonal antibodies (mabs), using affinity chromatography with protein A as ligand is very popular . . .
at a temperature in the range from about 10 °C. to about 18 °C.	The effect of temperature, 4°C versus ambient temperature (AT) (20-25°C), was studied for the mabs OT-hCG-1C, 4D, 3A, 6A and 7B and OT-HIV-4A and 4B . .
2. The method of claim 1 further comprising exposing the composition subjected to protein A affinity chromatography to a protease inhibitor	Whether or not degradation of the IgG molecule occurs, depends among other factors on pH and subclass of the mab. However, if required, the activity of cathepsin D can be inhibited by addition of pepstatin A
5. The method of claim 1 wherein the protein is an antibody.	The purification of immunoglobulins (IgG) . . .

ANDAs: TEVA DOES NOT INFRINGE NASONEX POLYMORPH PATENT

After filing of an ANDA by Teva for a generic version of Nasonex (mometasone furoate), Merck responded by filing a patent infringement suit based on U.S. Patent No. 6,127,353 (‘353 Patent).[1]  The ‘353 Patent claims a monohydrate crystalline form of the active ingredient in Nasonex.  Teva’s formulation was made with another crystalline form, but Merck argued that Teva’s crystalline form converted to the patented crystalline form over the two year shelf life of Teva’s product. 

On November 16, 2016, Judge Robinson of the U.S. District Court of Delaware found the '353 Patent to be valid, but not infringed by Teva.

The court first addressed the validity of the ‘353 Patent.  Teva asserted that the ‘353 Patent was invalid for double patenting based on the Federal Circuit’s holding in “Gilead,”  which held that a later-issued but earlier-expiring patent can serve as a double-patenting reference against an earlier issued but later-expiring patent.  Gilead Sciences, Inc. v. Natco Pharma Ltd., 753 F.3d 1208, 1212 (Fed. Cir. 2014).  Recently, the holding in “Gilead” was relied on to invalidate a patent covering the drug Remicade.[2]  In this case, a terminal disclaimer was required to revive another related application during prosecution (relating to the amount of time during which the application was abandoned, not to the subject matter of the claims). The parties disputed whether the patent that issued from the revived application qualified as a double patenting reference because it expired before the '353 patent.  The court held that the '353 patent was not invalid for double patenting since the patents-at-issue were from the same family, were examined by the same examiner at the PTO, and that “[t]his is not an instance of a patentee seeking to extend the patent term with ‘sequential’ applications.” 

After finding the patent to be valid,[3] the court found that Teva’s product did not infringe the ‘353 Patent.  The question for infringement was whether Teva's ANDA product (an aqueous suspension made with prior art anhydrous form) contained any patented monohydrate during the product's shelf life.  

Merck’s expert tested several of Teva’s batches after their expiration dates (2.5 years and 4 years after expiration).  Merck’s expert testified that Teva’s crystals changed to the patented crystals but did not know when these crystals formed, and "the testing did not tell us anything about whether [the patented monohydrate] was present before expiry." Instead, the testing of the expired samples only revealed that the patented crystals appeared at some point between when it was manufactured and when it was tested.  The court concluded that the expired samples were not representative of the ANDA product and that without testimony (or evidence) of when the monohydrate crystals formed in the expired products, the conclusory statements provided by Merck’s expert did not establish infringement.

Regarding Teva’s commercial batches, Merck’s expert allowed the crystals from Teva’s product to dry before performing single crystal X-ray diffraction (SCXRD) analysis.  Merck’s expert explained that he was not able to harvest the monohydrate crystals from a wet slide like he had from the samples of the expired batches.  (For the expired batches, Merck’s expert gave the bottle containing the product a small shake in order to disperse the suspension inside the spray bottle and sprayed a sample on a clean glass slide. He selected a particular crystal using optical microscopy; withdrew the crystal; mounted it onto a MiTeGen loop; and performed SCXRD on the crystal).  The parties disputed whether the drying of the slides promoted crystal growth.  Merck’s expert testified that he "viewed lots of slides where they were drying and ...noticed no formation of new crystals of any sort, including [the monohydrate]." He testified that "[d]rying itself doesn't provide a crystal. It's not part of our standard crystallographic practice. It doesn't happen."  Teva’s expert disagreed, testifying that when Merck’ expert allowed the wet slides to dry over extended period of time, he provided an uncontrolled experiment, which was actually conducive to formation of the monohydrate.

Merck’s expert testified that following a "learning period," he was confident in his ability to visually distinguish (with optical microscopy) between the monohydrate and anhydrous crystals. In summarizing his findings, he testified that he had identified "dozens and dozens" of monohydrate crystals in Teva’s products.  Teva’s expert disagreed with the reliance on visual observation alone, testifying that such observation "has to be coupled with X-ray crystallography of that same crystal in order to have any confidence of the chemical identity in the solid form of that crystal."

Merck also presented the testimony of Teva's 30(b)(6) deposition witness who testified that he could see a peak at 1710 cm^-1 in Raman spectroscopy (corresponding to a peak characteristic of the monohydrate).  The court, relying on case law,[4] concluded that at least three peaks on a spectra must be used to identify material based on accepted practices, and that Teva’s internal testing did not establish the presence of the monohydrate in Teva's product.

The court found that Teva did not infringe the ‘353 Patent since “the literature and the experts consistently pair optical microscopy with another measurement method before conclusively distinguishing polymorphs. . . the court finds that Merck has not established, by a preponderance of the evidence, the presence of [monohydrate] in Teva's ANDA product during its two-year shelf life.” At no point during his testing of Teva’s commercial batch did Merck’s expert harvest a monohydrate crystal from a wet slide, and only relied on crystals from slides which he had dried.

The court’s Order can be found at < https://www.docdroid.net/OqaCSNc/merck-opinion-nov-16-2016.pdf.html>.

---

[1] Merck Sharpe & Dohme Corp. v. Teva Pharmaceuticals USA Inc., case number 1:14-cv-00874, U.S. District Court for the District of Delaware.

[2] <https://www.pharmapatentsblog.com/2016/10/05/judge-grants-gilead-motion-to-invalidate-remicade-patent/>

[3] The District Court also dismissed Teva’s position that the ‘353 Patent lacked written description support

[4] Schering Corp. v. Apotex Inc., 2012 WL 2263292 (D.N.J. June 15,2012).