Why Novartis lost its treatment dispute with Teva

In a patent revocation and infringement counterclaim action between Teva and Novartis, His Honour Judge Richard Hacon found, in a decision handed down on November 10 2022, lack of inventive step with respect to two Novartis formulation patents ( EP 2,964,202 and EP 3,124,018)

He also ruled that Teva’s generic product ‘Teva DFX’ did not infringe the patents on either normal construction or on equivalents. The dispute involved a swallowable tablet version of Novartis’ ‘Exjade’ used to treat blood iron overload, previously only available as a dispersible tablet in Europe since 2006.

Novartis’s patents are obvious

The patents were previously upheld in the Opposition Division of the European Patent Office (EPO) with amended claims, with appeals pending. Novartis proposed to amend the patents to bring them into line with the claims allowed by the Opposition Division, which Judge Hacon agreed were allowable, and with each other.

The judge, however, found the patents to be obvious over two of the three prior art documents asserted by Teva on the basis that the invention was “an obvious modification” to each document, and that it made no difference that there was no incentive to make such modification.

Judge Hacon’s reasoning started with his analysis of the skilled team and the common general knowledge (CGK). The skilled team consisted of three individuals: a clinician, a formulator and a pharmacokineticist.

The CGK of the skilled team included the “attractive goal” of having swallowable deferasirox in place of the dispersible product “if it could be achieved”. The expectation before the priority date was that dispersible tablets had higher bioavailability in the patient when compared to a film-coated swallowable tablet because there was already a degree of disaggregation.

The patent specified the surprising result that the coated tablets of the invention had higher bioavailability than the dispersible form.

In dispersible form, the product had a long list of undesirable effects, special warnings and precautions for use. In addition, microcrystalline cellulose (MCC) was considered common filler and poloxamer 188 a common surfactant in tablet formulations and therefore part of the CGK.

Novartis’s evidence provided no interaction between its three experts, while Teva’s evidence sought to mimic real life by showing the interaction between its expert clinician and expert formulator. To demonstrate this, Teva produced a document crafted by its solicitors titled, ‘Skilled Clinician’s Instructions to Skilled Formulator’ with a wish list of what the clinician would want from the notional formulator.

The document was heavily criticised by Novartis, but the judge broadly agreed that its contents resembled what discussions between the skilled team at the priority date would look like, including the desirable improvements to dispersible deferasirox such as attempting a swallowable tablet, maximising dispersibility, and minimising grittiness.

Adopting Justice Birss’ comments in Hospira UK v Genentech (2014), the judge found that the steps which the skilled team would take at the priority date could readily be ascertained without the taint of hindsight, namely, that the use of deferasirox in swallowable film-coated tablet form and the selection of excipients MCC and poloxamer 188 would have been obvious to try with a reasonable expectation of success.

The judge made this finding despite the “limited commercial pressure to try” a reformulated deferasirox into a swallowable tablet.

Inventive concept cannot include matter not in the claims

Novartis argued the inventive concept was a film-coated tablet containing a high load of deferasirox with the claimed excipients, but without sodium lauryl sulphate and lactose, with higher bioavailability compared to ‘Exjade’. It further argued that it could be taken with food rather than on an empty stomach, the latter of which was only raised in closings. Teva’s inventive concept was simply a film-coated tablet formulation with an amount of 45-60% deferasirox.

Judge Hacon found that because bioavailability and reduced food effect were not set out in the claims of the patents, Novartis could not import these features into the claims or the inventive concept even if the skilled team reading the patent description would regard them as the most significant technical insight. The reason for this was that the inventive concept of a claim cannot encompass matter which forms no part of the invention as a whole.

The judge further found that the relevant question was whether it was obvious at the priority date to make a swallowable tablet as claimed and “nothing more than that”. Matter not within the claim is irrelevant to the assessment of the third and fourth Pozzoli steps.

If a product is obvious over a piece of prior art, the validity of a claim to that product cannot be salvaged by pointing to an unexpected “bonus effect” experienced when the product is used, even if it is the “clever bit” as Novartis argued.

Judge Hacon devised his own inventive concept as: “a swallowable film-coated tablet containing deferasirox in an amount between 45% and 60% of total weight and (subject to variation in conformity with the skilled person’s common general knowledge) containing (i) microcrystalline cellulose, (ii) crospovidone, (iii) povidone, (iv) poloxamer 188, (v) colloidal silicon dioxide and (vi) magnesium stearate, and containing no SLS or lactose.”

No infringement by Teva

Teva’s application for a declaration of non-infringement turned solely on the percentage content of deferasirox in Teva DFX, as outlined in its inventive concept. Novartis’s construction of “total weight”, which included the tablet coating, was favoured by the judge. However, as the amount of deferasirox in Teva DFX was, at its lowest, above the claimed range of 45-60%, Teva was outside of the patented claims on normal construction.

The judge considered that if Novartis’s version of the inventive concept were correct, Teva DFX would infringe as an equivalent. On the basis of the judge’s inventive concept, and having considered the Actavis questions, the variant could not be substantially the same as the inventive concept unless it strictly complied, and thus there was no finding of equivalence.

Azadeh Vahdat is an associate at EIP. She can be reached at avahdat@eip.com