In In re Entresto (Sacubitril/Valsartan) Patent Litigation, Judge Richard G. Andrews of the U.S. District Court for the District of Delaware granted MSN Pharmaceuticals Inc. a victory on noninfringement of U.S. Patent No. 11,096,918 (’918 Patent), finding Novartis’ scientific evidence to be unreliable, and accordingly drew an adverse inference against Novartis because of its failure to produce samples underlying its expert’s infringement analysis. Following protracted litigation involving multiple patents and defendants, MSN’s victory potentially opens Novartis’ very successful branded product, with billions in sales, to immediate generic competition.
Background
Novartis’s new drug application (NDA) for Entresto, a combination therapy of valsartan and sacubitril, was approved in 2015 and indicated “to reduce the risk of cardiovascular death and hospitalization for heart failure in adult patients with chronic heart failure, and for the treatment of symptomatic heart failure with systemic left ventricular systolic dysfunction in pediatric patients aged one year and older.”
The ’918 Patent was not listed in the Orange Book in conjunction with the NDA and, therefore, did not trigger an automatic statutory stay barring the Food and Drug Administration from approving abbreviated new drug applications (ANDAs) for generic versions of the branded product. As a result, MSN does not face any barrier to launch upon FDA approval following expiration of a regulatory stay on a different patent that is no longer in the litigation.
The relevant claims of the ’918 Patent were directed to an amorphous solid form of trisodium valsartan sacubitril (TVS):
- An amorphous solid form of a compound comprising anionic [valsartan], anionic [sacubitril], and sodium cations in a 1:1:3 molar ratio.
In 2019, MSN and other manufacturers filed ANDAs, and Novartis sued under 35 U.S.C. § 271(e)(2), alleging that the generic products would infringe the ’918 Patent because they would contain amorphous TVS.
At claim construction, the parties disputed the meaning of a single claim term: “an amorphous solid form of a compound.” Novartis took the position that any presence of amorphous TVS in a solid form would fall within the scope of the claims, while MSN argued that the claimed amorphous compound must be “substantially pure.” The court rejected both constructions but agreed with MSN’s position that “construction of the disputed term must distinguish between the amorphous and crystalline forms.” The disputed term “an amorphous solid form of a compound” was construed to mean:
[A] solid form of a compound in which the amorphous form of the compound predominates. An amorphous solid form is mutually exclusive from a crystalline solid form, but not necessarily mutually exclusive from a partially crystalline form.
Unreliable Expert Testimony
Novartis relied on testing done by Raman spectroscopy that could distinguish between a physical mixture of multiple compounds and an amorphous complex of the same multiple compounds. The parties agreed that the Raman spectrum would differ for physical mixtures versus amorphous complexes.
Novartis’ expert created a spectrum of an amorphous physical mixture by generating separate spectra for amorphous valsartan disodium and amorphous sacubitril sodium, normalizing the two spectra, then adding them together. The court found in favor of Novartis on the reliability of this mathematically generated Raman spectrum as representative of an amorphous physical mixture.
However, Novartis failed to establish that its expert successfully distinguished between amorphous TVS and an amorphous physical mixture in the Raman spectrum analysis.
Novartis claimed that a reference “glassy solid” created by its expert was amorphous TVS rather than an amorphous physical mixture. Novartis’ expert testified that she compared the spectra of the glassy solid to those of an amorphous physical mixture and found the two to differ. Novartis sought to establish the difference between the two spectra by zooming in on a subsection of overlaid spectra, showing a shift in peaks between the spectra. Novartis argued that this peak shift established that the glassy solid was amorphous TVS with a Raman spectrum distinguishable from that of amorphous mixture.
MSN contended that the spectrum when viewed as a whole, rather than in an isolated section, showed that the entire Raman spectrum of the glassy solid shifted approximately 3 cm-1 to the right of the Raman spectrum of the physical mixture. The court gave credence to MSN’s expert’s argument that the type of systematic shift observed between the Raman spectrum of the amorphous physical mixture and the glassy solid would be “impossible” if the glassy solid were indeed an amorphous complex, which would necessarily have a different pattern of peak positions and intensities. The court found support for its position in a Novartis reference explaining that differences in peak positions and intensities could be used to differentiate between an amorphous complex and a physical mixture but was silent on the type of systematic shift seen in Novartis’ spectrum.
The court also pointed to Novartis’ own internal testing of a compound termed amorphous LCZ-696, allegedly the amorphous complex, which showed a Raman spectrum different from glassy solid Raman spectrum generated by Novartis’ expert for the purposes of the litigation. While there was no definitive testimony establishing that LCZ-696 was amorphous TVS, Novartis offered no testimony confirming that amorphous LCZ-696 was in fact not amorphous TVS. Based on the evidence before it, the court found there was a reasonable likelihood that amorphous LCZ-696 was amorphous TVS. Since Novartis’ internal data showed that the amorphous LCZ-696 spectrum differed from the glassy solid spectrum, this cast further doubt on whether Novartis’ glassy solid was a reliable reference standard for comparison with MSN’s ANDA.
Adverse Inference
MSN’s codefendants Gerbera Therapeutics Inc. and Nanjing Noratech Pharmaceutical Co. Ltd. (together, “Noratech”) had requested samples of the glassy solid, which Novartis failed to produce. At the pretrial conference, the court indicated that it would draw an adverse inference against Novartis for failing to produce the sample.
Although Noratech had settled and MSN had not separately requested the samples, the court dismissed Novartis’ argument that no adverse inference was warranted because MSN was not the party requesting the samples and had failed to pursue a motion to compel. The court noted that MSN had been precluded from testing the samples and that Novartis did not claim that it was not in possession of the samples. The court concluded that the samples had been withheld because “had Novartis produced the glassy solid sample to Defendants, it would have been unfavorable to Novartis’ case.”
Noninfringement Finding
The court found that Novartis failed to establish infringement by a preponderance of the evidence because its expert’s analysis was unreliable. In addition, the adverse inference arising from Novartis’ failure to provide a glassy solid sample supported the conclusion that “Novartis did not meet its burden to show MSN’s ANDA infringes by a preponderance of the evidence.”
Conclusions
MSN’s victory hinged on multiple factors, including a favorable claim construction that obligated Novartis to establish that MSN’s ANDA product contained substantial amounts of the amorphous complex, a finding that Novartis’ testing was unreliable, and an adverse inference against Novartis for failing to provide the samples developed by its expert for testing. While Novartis is expected to appeal this ruling promptly to prevent MSN from launching its competing product during the pendency of the appeal, MSN’s win serves as a reminder of the importance of a favorable claim construction outcome and the importance of reliable and credible expert testimony.
