UK court finds patent invalid for lack of plausibility across the scope of the claims

On 24 July 2023, Meade J handed down his decision in Teva v Grünenthal [2023] EWHC 1836 (Pat) concerning EP (UK) 1 457 208 (the “Patent”) which relates to a formulation of testosterone undecanoate (“TU”) with castor oil and benzyl benzoate for treating hypogonadism. The Judge dismissed the obviousness attack, but found the Patent to be insufficient for lack of plausibility and therefore invalid.

The relevant product claims in issue concerned a TU formulation in a vehicle comprising 25-45% castor oil and an unspecified amount of benzyl benzoate (Claim 1), and a claim dependent on Claim 1 comprising 55-65% benzyl benzoate (Claim 4) (the “Product Claims”). The presence of an oil and co- solvents in the vehicle is essential for achieving a “depot effect”, i.e. a slow release of the drug into the body which meant that patients could have longer intervals between injections whilst keeping their testosterone levels in the normal range. The relevant method claims in issue claim a method of administrating the TU formulation using 2 loading doses at 6-weekly intervals followed by a maintenance phase of injections in intervals of at least 9 weeks (Claim 7), or the same loading doses following by injections at 10 or 12 weekly intervals (Claim 16) (the “Method Claims”).

Grünenthal markets a formulation of TU covered by the Patent under the brand name Nebido®.

Obviousness over the prior art
Teva pleaded obviousness over a piece of prior art referred to as “von Eckardstein”, read in light of two papers referenced therein (“Nieschlag” and “Behre”), all of which related to a series of clinical studies for a TU in castor oil formulation for treating male hypogonadism. It was proven at trial that the TU formulation in the prior art was the same as Nebido.

The first issue concerned the Product Claims, and whether the disclosure in von Eckardstein of the TU preparation being “1000mg TU dissolved in 4 ml castor oil” was a complete disclosure, or whether it included an unspecified co-solvent and if a co-solvent were included, the identity of the co-solvent and the amount. Teva’s case was that the skilled person would know that castor oil was far too viscous to be injected into people on its own and therefore a co-solvent was necessarily included in the formulation.

Teva argued, based on the commercially available formulations at the priority date, that benzyl benzoate was the most obvious partner for castor oil and from then on, routine tests would reveal the ideal ratio of castor oil to benzyl benzoate in order to yield an optimal viscosity for the formulation whilst ensuring the TU remained dissolved. It argued that the skilled person would reverse-engineer a formulation which had a similar overall profile to a dose response curve in the cited prior art.

Meade J accepted that upon testing the solubility of TU in castor oil the skilled team would quickly discover there must be a co-solvent present and that benzyl benzoate is one option the skilled formulator would try. However, he considered that Teva’s argument that the skilled person would seek to reverse-engineer the prior art was driven by impermissible hindsight.

The second issue concerned the Method Claims, and whether it would be obvious for the skilled team to alter the dosing regimen disclosed in the prior art to arrive at the Patent. Teva’s case was that the skilled team would understand the prior art to disclose 4 loading doses, which achieved supraphysiological levels of serum testosterone after the 3rd and 4th injections. Therefore it would be obvious to reduce the number of loading doses down to two. The Judge again rejected this argument as being tainted with hindsight; the skilled team would not know that two loading doses were better than four, and there were other posology options the skilled team would try.

Insufficiency
Considering insufficiency, the Judge applied the three-step test from Fibrogen which, at least until the Supreme Court determines otherwise, represents the approach the determining the sufficiency of claims to a class of compounds:

1. What falls within the scope of the claimed class?;
2. What does it mean to say that the invention works?;
3. Is it possible to make a reasonable prediction that the invention will work with substantially
everything falling within the scope of the claim?

There was no dispute on question 1. On question 2, Meade J referred to the specification to find that the invention must provide physiological levels of testosterone for a prolonged period of time (and achieving the claimed dosing intervals for the Method Claims).

The real dispute turned on question 3. The Judge accepted Teva’s submission that the claims (as interpreted by the teaching of the specification) allowed for substantial variation in parameters such as the amount of TU injected, and the percentages of castor oil, benzyl benzoate and other excipients in the vehicle, and found that variations in these parameters would have knock-on effects on solubility, viscosity, partition coefficient and ultimately on the depot effect. All the claims in issue were therefore insufficient for lack of plausibility.

Finally, Meade J considered the decision of the German Federal Patent Court which had revoked the German designation of the Patent for obviousness (lack of plausibility was not in issue) earlier in the year. The Judge concluded that Teva’s obviousness attack and the evidence adduced in the German proceedings was quite different to the UK case, therefore he was able to reach a different conclusion.

A copy of the judgment can be read here.

Headnote and summary: Milly Wickson, Bristows LLP