Patents – pharmaceutical products – paracetamol tablets – erroneous description of testing apparatus –whether claims should be construed so as to avoid the error – level of detail required in describing ingredients in pharmaceutical formulations – relevance of manufacturing and processing information in marketing authorisation application

GlaxoSmithKline Consumer Healthcare Investments (Ireland) (No 2) Ltd v Generic Partners Pty Ltd[2018] FCAFC 71

Pain is, well, a pain. One medication commonly used to relieve and manage pain and fever is paracetamol. Paracetamol is safe in low doses, but at high doses can cause liver damage and other problems. For this reason, the usual dosing regimen of paracetamol tablets is two tablets (each containing 500mg of paracetamol) every 4 to 6 hours.

When a drug like paracetamol is administered, it is adsorbed by the body, distributed through blood plasma and body tissues, broken down (or “metabolised”) and eventually eliminated; the overall process of a drug moving through a living system in this way is referred to as “pharmacokinetics”.

These pharmacokinetic processes mean that the concentration of a drug administered to a living system will change over time. Because of this, ensuring that the amount of paracetamol present in a person’s system remains both sufficient to provide effective pain relief (that is, the drug’s “therapeutic level”) and safe can be challenging. In other words, in safe doses a paracetamol tablet may start to “wear off” before it is time to take another dose.

One solution is to formulate paracetamol tablets with two layers, one that dissolves quickly and releases an immediate dose of paracetamol (“immediate release”) and another that dissolves more slowly and so releases its dose over a longer period (“sustained release”). McNeil Inc’s Tylenol® Extended Relief is an example of such a “bilayer” paracetamol tablet, and is the subject of a patent that discloses a tablet containing between 650 and 667mg of paracetamol, distributed equally between immediate release and sustained release layers (“McNeil tablet”). It should also be noted that such a tablet contains more paracetamol than is present in standard paracetamol tablets.

The appellants (collectively “GSK”) have rights to a patent for another bilayer paracetamol tablet (“GSK patent”). The GSK patent discloses a bilayer tablet containing between 600 and 700mg of paracetamol, but with a different allocation of paracetamol between the immediate release and sustained release layers. Claim 1 of the patent specifies constraints for the tablet’s “dissolution profile” (that is, a drug’s concentration as a function of time after being added to a dissolution medium, or its “release rate”), to be determined by a “USP type III apparatus, reciprocating basket” under particular conditions. Tablets with a dissolution profile falling within the specified constraints are said to be superior to the McNeil tablet because they have a lower “fluctuation index” or “FI” (a measure of the variability in concentration of a drug over time); the lower the FI, the lower the risk of accidental overdose and the safer the product.

The dissolution profile of a drug is tested in vitro using an apparatus under conditions that try to mirror the pharmacokinetics of the human body. A “USP type III” apparatus is one of four apparatuses used to measure a drug’s dissolution profile; the others are referred to USP type I, type II and type IV. It is necessary to identify the apparatus with specificity, because the different types of apparatus use different methods to move fluids involved in the testing, and this affects the dissolution results. Testing the same drug using different apparatuses will produce different dissolution profiles, and there is no established correlation between those results.

The reference in claim 1 to a “type III apparatus, reciprocating basket” is incongruous, because while a type I apparatus uses a basket, a type III apparatus uses a reciprocating cylinder. As at the priority date of the GSK patent, 19 April 2000, it theoretically was possible to use a type I basket instead of the cylinder of a type III apparatus, but there would be no guarantee the apparatus would work, or produce the same dissolution profiles as the standard type III apparatus. Nevertheless, the reference to a “basket” being used with a type III apparatus would not have been dismissed by a skilled address as being ridiculous nor turned the invention into a basket case.

At trial, a key issue was how a skilled addressee would construe the reference to a “reciprocating basket” in the context of a “type III apparatus” described in the claims of the GSK patent, in light of the common general knowledge at the priority date. The learned trial judge accepted that the skilled addressee would understand that “basket means basket” and that the reference to “basket” was erroneous, but contrary to GSK’s submissions declined to construe claim 1 as referring to a type III apparatus with a reciprocating cylinder; his Honour considered such an approach to involve an impermissible re-writing of the claim “under the guise of construction”. His Honour also held that the hypothetical type III apparatus using a basket would have been considered to be possible, and further would “still work”. His Honour’s findings were fatal to GSK’s infringement action against the respondents (for convenience, “Apotex”).

On appeal by GSK, the Full Court (Middleton, Nicholas and Burley JJ) found no error with the trial judge’s findings of fact referred to above, and accepted that a skilled addressee would understand the reference to a “reciprocating basket” to be an error. As the Full Court observed at [112]:

[the] question is whether, if the court were to also interpret the claim so as to correct the mistake, it would be re-writing the claim or merely interpreting it through the eyes of the skilled addressee in accordance with the principles to which we have referred. [Emphasis added.]

The Full Court noted that it is for the court to construe the meaning of claims; the skilled addressee’s understanding of claim 1 cannot override the function of claims in defining an invention. In this case, either ignoring the words “reciprocating basket” (and thereby reverting to the standard set up of a type III apparatus) or interpreting “basket” to mean “cylinder” effectively would re-write the claim. Accordingly, claim 1 properly was to be construed as referring to a type III apparatus with a reciprocating basket, and GSK’s appeals were dismissed.

At trial, Apotex was unsuccessful on a cross-claim to have the GSK patent revoked for lack of fair basis and failure to disclose the best method known to GSK for performing the invention. Apotex cross-appealed on those issues.

As to fair basis, Apotex argued that the specification in the GSK patent disclosed a single formulation for a bilayer paracetamol tablet with a “unique” dissolution profile, and that that invention was narrower than the scope of the invention described in the consistory clauses of the specification. Apotex submitted that in rejecting Apotex’s argument, the trial judge erred by going beyond a pure textual analysis of the specification, looking further to matters relating to common general knowledge (like expert evidence and extraneous documents) and to questions of reasonableness, utility and sufficiency.

The Full Court rejected Apotex’s criticisms of the trial judge’s approach, and the proposition that the specification describes an invention narrower than that described in the consistory clauses; the skilled addressee (equipped with the common general knowledge) would understand the specification to refer to tablets with dissolution profiles falling within a range of possibilities. The Full Court also noted that a specification does not need to explain the parameters of that range, or otherwise describe the invention in a different or more detailed way than the claims; while a complete specification must “describe the invention fully”, it is not necessary that the specification explain why the invention works.

On best method, the Full Court accepted that a particular formulation described in the specification represented the best known method for formulating the tablet. Apotex’s attack was founded on one ingredient of that formulation, a polymer powder referred to as “HPMC”, being identified only as “high viscosity HPMC”, without specifying the particular grade or viscosity or what degree of granulation prior to compression of the tablet was used. However, in its marketing authorisation application (“MAA”) for the drug it in fact marketed (completed just prior to the filing of the application for the GSK patent), GSK had included specific information about the grade and viscosity of the HPMC, and the degree of granulation, used which led to the tablet being as close to the “ideal” dissolution profile as possible. The Full Court noted that that fact did not necessarily mean that the MAA disclosed the best method, as opposed to some commercially expedient method. Further, the expert witnesses did not know whether granularity would have a significant impact on dissolution profiles. Accordingly, the Full Court (at [203]) rejected Apotex’s argument and upheld the trial judge’s finding that the best method had been disclosed, “albeit at a level of generality that did not include the more detailed but inessential manufacturing and production information described in the MAA”.

In short, none of the parties were successful at trial or on appeal; this was, no doubt, a bitter pill to swallow.