Bausch Health Cos. v. Actavis Labs. FL

Case Name: Bausch Health Cos. v. Actavis Labs. FL, Civ. No. 16-9038 (SRC), 2019 U.S. Dist. LEXIS 118876 (D.N.J. July 17, 2019) (Chesler, J.)

Drug Product and Patent(s)-in-Suit: Relistor® (methylnaltrexone bromide); U.S. Patent No. 8,524,276 (the ’276 patent)

Nature of Case and Issue(s) Presented: Plaintiffs own the ’276 patent, which claims the oral methylnaltrexone bromide tablet formulation sold as the constipation drug Relistor. Actavis filed an ANDA for its own methylnaltrexone bromide tablet, and plaintiffs filed suit. Two issues were presented at trial: (i) whether defendants infringed claim 2; and (ii) whether claims 2 and 5 were invalid as obvious. After a four-day bench trial, the court found that Actavis infringed claim 2 and that claims 2 and 5 were not obvious.

Why Plaintiffs Prevailed: Claim 2 of the ’276 patent requires that sodium dodecyl sulfate (SDS) form an ion pair with methylnaltrexone, or a pharmaceutically acceptable salt thereof. The parties stipulated that SDS is also known as sodium lauryl sulfate (SLS). Plaintiffs argued that the combination of methylnaltrexone and SLS, as the ’276 patent teaches, results in an increased apparent octanol/water partition coefficient (APC) of the methylnaltrexone. This increased APC allows for greater permeability of the constipation drug. Plaintiffs’ expert tested the Actavis ANDA product samples and observed an increased APC. Plaintiffs argued the increased APC of Actavis’s ANDA product evinced ion pairing between methylnaltrexone and SLS as required by claim 2.

In rebuttal, Actavis argued that the APC test run by plaintiffs’ experts was unreliable and unprecedented for determining whether two substances in a tablet from an ion pair in solution. That expert admitted he could not determine whether the increased APC was due to ion pairing or a different interaction. Further, Actavis argued the APC test could not determine whether two substances in a tablet form an ion pair in solution. Another of Plaintiffs’ experts responded that an increase in APC is the standard way to show ion pairing.

The court found that Actavis’s ANDA product combined methylnaltrexone and SLS to form an ion pair as required by claim 2. The court dismissed Actavis’s argument that an increase in APC does not necessarily demonstrate ion pairing. Rather, the court gave significant weight to plaintiffs’ experimental evidence that the APC of methylnaltrexone was significantly greater in Actavis’s ANDA product than a product without SLS. The court also disagreed with Actavis about whether APCs can be determined for ingredients within tablets. The court found that the ’276 patent requires APC determination and rejected all of Actavis’s challenges to plaintiffs’ infringement case with one exception: the general caution that questions remain around the change in the APC of methylnaltrexone where combined with multiple excipients. Examining plaintiffs’ experiments, the court found that ion pairing of methylnaltrexone and SLS caused the increase in the APC of the ANDA product. Thus, plaintiffs proved infringement of claim 2 (the parties stipulated to infringement of claim 5).

Turning to Actavis’s obviousness arguments, claim 5 teaches that at least 50% of the tablet comprising methylnaltrexone, or a pharmaceutically acceptable salt thereof, and SDS dissolves in a dissolution apparatus. Actavis argued that oral formulations containing methylnaltrexone were known in the art at the time FDA approved Relistor, as were the permeability issues surrounding methylnaltrexone. Actavis argued a POSA would have been motivated to develop an oral formulation of methylnaltrexone and would have been motivated to utilize a known permeation enhancer, the excipient SLS, to address those issues. Plaintiffs responded that the prior art recognized that ion pairing was uncommon and controversial, and any ion pairing between methylnaltrexone and SLS could only be known via testing. Further, Plaintiffs showed that increasing partitioning, as shown in prior art, does not necessarily increase permeability. Finally, plaintiffs explained that the formulation of methylnaltrexone and SLS dramatically improved efficacy, yielding much faster clinical laxation results than tablet formulation without SLS. Actavis argued this could not be evidence of objective indicia of non-obviousness because plaintiffs had not shown it to be unexpected as compared to the closest prior art.

In analyzing the obviousness question, the court first disagreed that methylnaltrexone had “poor” permeability. Rather, the court recognized that while the prior art shows methylnaltrexone may have lower lipid-solubility than other antagonists, it does not support that its solubility is deficient or poor. The court found that low bioavailability of methylnaltrexone, as shown by the prior art relating to plasma bioavailability, provided no evidence as to its permeability. Instead, the court determined that hindsight guided Actavis’s theory.

The court also took issue with Actavis’s argument that SLS would have been a POSA’s clear choice to improve permeability and bioavailability of methylnaltrexone. First, the proposition lacked the factual support that methylnaltrexone had poor permeability. Second, Actavis’s expert relied on prior art categorizing, identifying, and formulating approaches for oral bioavailability problems. The court found the prior art taught away from increasing bioavailability. Further, while one of the mentioned “bioavailability problems” was “poor membrane permeation,” a POSA would not even have sought this prior art, because methylnaltrexone was not known to have bioavailability problems. The court similarly found that, because a POSA would not have recognized methylnaltrexone as having poor permeability, it would not be obvious to select SLS in particular as a permeation enhancer. The court also credited plaintiffs’ argument that the prior art taught away from combining SLS with methylnaltrexone.