Actelion's Patent Claimed "pH 13." It Never Said At What Temperature - And That Silence Just Cost It the Case

Epoprostenol is a genuinely difficult molecule to keep alive in a vial. It's a vasodilator used to treat pulmonary hypertension, and it reacts with water almost as fast as you can dissolve it in it - hydrogen ions in solution catalyze its breakdown into a compound with none of its pharmacological value. The older product, Flolan, worked around this with a special basic diluent and mandatory refrigeration. Actelion's patents on its successor drug, Veletri, promised something more convenient: a formulation stable enough to survive reconstitution with ordinary IV fluid at room temperature.

The chemistry behind that promise is straightforward acid-base control. Push the pH of the "bulk solution" - the mixture before it's freeze-dried into powder - high enough with arginine and sodium hydroxide, and you starve the degradation reaction of the hydrogen ions it needs. Actelion's data showed pH 13 was the real threshold: below it, the powder and the reconstituted solution both degraded meaningfully faster. That number went straight into the claims. Every independent claim requires the lyophilized composition be "formed from a bulk solution having a pH of 13 or higher."

Here is the part that unraveled fifteen years later, in Actelion Pharmaceuticals Ltd v. Mylan Pharmaceuticals Inc. (Fed. Cir. 2024-1641, decided May 13, 2026): pH is not a fixed property of a substance. It's a measurement, and measurements need conditions. The exact same solution reads differently on a pH meter depending on temperature - water's own autoionization shifts with heat, which is why "neutral pH 7" is itself only true at 25 degrees Celsius. Actelion's patent used "pH 13" more than a dozen times across its specification and claims. It never once said at what temperature that number was supposed to hold.

The Federal Circuit didn't buy it, and its reasoning is a genuinely instructive read for anyone drafting or fighting a claim built around a measured physical property. The court didn't reach for the claim language itself - it admits upfront that "a pH of 13 or higher" simply doesn't specify measurement conditions at all, so the words alone couldn't settle it either way. Instead it built the answer almost entirely from the specification's own internal consistency. Every experimental table in the patent reports pH values without ever adjusting for temperature; the spec defines "alkaline" as pH greater than 7, a fact only true at standard temperature; and critically, the parties' own experts agreed that every single pH-13 data point Actelion generated to prove its invention worked was measured the ordinary way; at room temperature. A patentee cannot prove unexpected results at standard-temperature pH 13, then turn around and argue infringement should be judged at a colder, more favorable number for the accused product. The extrinsic evidence sealed it further: the United States Pharmacopeia, the industry's own reference standard, states plainly that pH values mean standard temperature "unless otherwise specified." Actelion's patent never specified otherwise.

That alone was enough to end the literal infringement claim. What makes this case worth sitting with, though, is that Actelion lost the doctrine-of-equivalents fallback twice over, on two completely independent grounds, and both traces back to decisions made years earlier during prosecution.

The second, independent bar is the disclosure-dedication rule, and it is the sharper lesson for patent drafters specifically. Actelion's own specification states the bulk solution pH is "preferably adjusted to about 12.5-13.5, most preferably 13" and, separately, "preferably greater than 11, preferably greater than 12, and, most preferably greater than 13." Both sentences disclose pH ranges below 13 as viable alternatives - and then the claims simply never captured them. Under settled Federal Circuit doctrine going back to Johnson & Johnston, subject matter you disclose but don't claim is dedicated to the public. You cannot get it back later by calling it an "equivalent." Actelion tried arguing this shouldn't apply because the disclosed ranges overlapped rather than being mutually exclusive alternatives - the court rejected that outright, noting nothing in the doctrine requires disclosed alternatives to be tidy and non-overlapping to count.

Two independent legal doctrines, both triggered by choices Actelion's own patent drafters made a decade before this litigation started, both closing the exact same door. For a generic manufacturer weighing whether a formulation patent built around a measured physical parameter is worth challenging, the real takeaway isn't "argue about temperature" - it's this: go looking for what the specification discloses but the claims don't reach, and go looking at exactly why the claims were narrowed during prosecution. Both are permanent, and neither depends on winning the underlying science argument at trial.