Full Federal Court hands down a bitter pill for Sanofi

Full Federal Court hands down a bitter pill for Sanofi

Full Federal Court hands down a bitter pill for Sanofi

Apotex Pty Ltd v Sanofi-Aventis [2009] FCAFC 134

The recent decision of H Lundbeck A/S v Alphapharm Pty Ltd (see our IP Update, September 2009 issue), affirmed that individual enantiomers which make up a racemate can be patentable in light of a prior disclosure of the corresponding racemate. However, in the present case, the Full Court has held that prior disclosure of the racemate accompanied by a general reference to the individual enantiomers is sufficient to anticipate subsequent claims to one particular enantiomer.

Facts of the case

In the early 1980s Sanofi Aventis (Sanofi) obtained patents in Australia, France, Canada and the United States in respect of a series of thienopyridine compounds for inhibiting action on blood aggregation and anti-thrombotic activity. The compounds possess a chiral centre and thus afford two possible enantiomers, which may be designated as the (d)- (or ( )-) and (l)- (or (-)-) forms, depending on the direction in which they rotate plane polarised light. Each illustrated the synthesis of a particular racemic compound (a 1:1 mixture of each enantiomer, resulting in no net rotation) referred to as PCR 4099. Subsequent testing of PCR4099 indicated adverse side effects and attempts to resolve and test the individual enantiomers were undertaken. This resulted in the discovery that the (l)- enantiomer was inactive and that the activity of PCR4099 lay in the (d)-enantiomer, and that furthermore, the (l)-enantiomer was more than twice as toxic as the (d)- enantiomer. The (d)-enantiomer (known generically as clopidogrel) was launched in Australia in 1998 as its bisulphate salt under the name Plavix and is the subject of Australian Patent No 597784 (the patent). Claim 1 relates to the (d)-enantiomer and its pharmaceutically acceptable salts, claims 2- 5 are to specific salts of the (d)-enantiomer, claims 6-9 are to processes for preparing a compound of claim 1, and claims 10 and 11 are to compositions comprising a compound of claim 1. Under the pharmaceutical extension of term provisions of the Australian Patents Act 1990, the patent term was extended to 2013.

Apotex Pty Ltd (Apotex) sought revocation of the patent on the grounds of, inter alia, lack of novelty and inventive step. As reported in our IP Update December 2008 issue, the primary Judge found that claims 1, 10 and 11 lacked novelty and claims 6-9 were invalid for want of an inventive step. Claims 2-5, being directed to a number of specific pharmaceutical salts, were, however, found to be valid. Sanofi appealed from the finding that claims 1, and 6-11 were invalid and Apotex appealed from the finding that claims 2-5 were valid.

The decision



In considering the issues, the Court construed the claims and prior art in the position of the skilled addressee and began by examining the contents of the Australian, French, US and Canadian prior art patents. Each set out the general formula for the thienopyridines and a process for their preparation, as well as describing their therapeutic applications as blood-platelet aggregation inhibitors. Specifically, the patents stated that “The invention relates to both each enantiomer and their mixture“ and further claimed “one of the 2 enantiomers or their mixture“. A process for the preparation of PCR 4099 was described but no guidance was given for the preparation of the enantiomers referred to, although the Canadian patent did note that the enantiomers could, if desired, be separated. It was, however, accepted that the separation of the enantiomers was a routine matter and involved no inventive step. On this basis, the Court unanimously agreed with the primary judge and held that the prior art patents unequivocally described and claimed the enantiomers, predicted to have the beneficial qualities of the racemic compounds exemplified, and that there was a clear direction to the skilled reader to prepare the enantiomers.

Whilst the primary judge had examined the concept of selection patents, the Full Court did not consider it was necessary to decide on this point. Its view was that any such category would not exclude from the requirement of novelty a compound that was previously disclosed and claimed as one of a class of compounds that demonstrated or were predicted to demonstrate a particular activity and tolerance that was subsequently shown to demonstrate that activity at a high level and tolerance. Accordingly, the (d)-enantiomer, and therefore Claims 1, 10 and 11 were held to lack novelty in light of the prior art patents.

Claim 1 of the patent related not only to the (d)-enantiomer but also to its pharmaceutically acceptable salts, with claims 2-5 respectively reciting hydrochloride, hydrogen sulphate, hydrobromide and taruocholate salts. The Court examined the prior patents and concluded that although they made reference to and exemplified salts with pharmaceutically acceptable mineral or organic acids, the exemplified salts were in respect of the racemic compounds and the language used in the prior art patents was such that this was construed to be in the context of the racemate and not the individual enantiomers. Accordingly, the majority (Middleton and Bennett JJ, with Emmett J dissenting) concluded that the pharmaceutically acceptable salts of the (d)- enantiomer were novel over the prior art patents. The majority further noted that although three of the specified salts in claims 2-5 were described in the prior art patents in the context of a racemate, the prior art did not contain a clear description of, or instructions to make, the specific salts of the (d)-enantiomer of PCR4099 and that claims 2-5 were therefore also novel over the prior art patents.

Inventive Step

The provisions of the Patents Act 1952 governed the assessment of inventive step. Inventive step could therefore only be determined by reference to what was the common general knowledge in Australia at the relevant priority date. From the evidence presented to the primary judge, which was not disputed, it was accepted that a range of straight forward and routine techniques for obtaining enantiomers from a racemate, including the use of diastereomeric salts, was well known prior to the priority date, and that the formation of pharmaceutically acceptable salts of both racemates and enantiomers was a common process. Furthermore, it was also accepted as being well understood prior to the priority date, that enantiomers can exhibit different biological activity and that most or all the activity may lie with one enantiomer.

The Court agreed with the primary judge that for claim 1 the asserted inventive step related not so much to the desirable characteristics of the (d)-enantiomer, for which the activity was in the expected range, but to the lack of activity of the (l)-enantiomer. However, in light of the common general knowledge, this discovery was not considered to be so unexpected so as to amount to an inventive step, and that although there was a choice to be made in terms of the particular pharmaceutical salts, the salts specified in claims 2-5 were all conventional and the skilled addressee would know to make a salt that was pharmaceutically acceptable. Claims 2-5 were therefore held to lack an inventive step.

With regard to the process claims 6-9 for obtaining the (d)-enantiomer by diastereomeric salt formation, the Full Court also held that the primary judge was not in error in his finding that these were well known and obvious processes, and accordingly, claims 6-9 were also found invalid for lack of an inventive step.


When considered along side the recent decision of H Lundbeck A/S v Alphapharm Pty Ltd, in which an individual enantiomer was found patentable, (see our IP Update, September 2009 issue), it is clear that the patentability of a single enantiomer in light of the prior disclosure of the racemate is dependent on the particular facts. In the Lundbeck decision, a prior art patent disclosed the racemate but made no reference to or suggestion of individual enantiomers. Moreover, in Lundbeck the process for resolving the enantiomers was not a simple routine matter but was only achieved after many years of attempts. In light of the present findings, it is evident that for patentability of a single enantiomer, something more is required than simply resolving individual enantiomers in a routine manner and determining which enantiomer possesses activity and/or toxicity, particularly where the prior art provides the express motivation to do so.

An application for special leave to appeal has been filed with the High Court.