Extended Release SEROQUEL is Obvious To Try

AstraZeneca Canada Inc v Teva Canada Ltd / quetiapine ER (NOC) 2013 FC 245, 2013 FC 246 Near J
            SEROQUEL XR / quietiapine fumarate extended release tablets / 2,251,944

Near J’s decision in this NOC proceeding is a good illustration of the application of the obvious-to-try analysis to invalidate a patent.

The patent claimed a sustained release formulation of quetiapine fumarate using the gelling agent hydroxypropyl methylcellulose (HPMC). Quetiapine was a known drug used for the control of psychiatric disorders such as schizophrenia [3]. It was known that less frequent dosing is desirable as it leads to better compliance [30]. HPMC was a known gelling agent; indeed, it was one of the most popular gelling agents for use in extended release formulations [31]. The question was whether it was obvious to try using quietiapine in combination with HPMC to create a sustained release formulation [34].

AstraZeneca argued that as a matter of law, in order for the obvious-to-try test to apply, “it must be obvious that successful results will be achieved before any experimentation is carried out” [36]. Near J rejected this position (at least on the facts of this case) [37]. In my view, he was entirely correct to do so. The whole point of the obvious-to-try analysis is that advance prediction of results may be impossible “in areas of endeavour where advances are often won by experimentation” Sanofi 2008 SCC 61 [68], but if it is obvious to try the experiment, which is routine and successful, the invention will nonetheless be obvious. AstraZeneca also argued that there were reasons why it was not obvious to try making a sustained release formulation of quietiapine at all, and also why HPMC was not an obvious choice as a gelling agent [51]. Near J rejected these arguments on the facts. Accepting his factual conclusions, this is about as clear a case as I can imagine of an invention being obvious to try.

There are a couple of ambiguities in Near J’s description of the legal test. Teva’s position, which he held to be “more apt on the facts of this case” [37], was that “a patent will be obvious if it was more or less self-evident, in the words of Sanofi, to ‘try to obtain the invention’ or, in Teva’s words, to conduct routine experimentation with a fair expectation of success” [36]. This suggests that “more or less self-evident” is equivalent to a “fair expectation of success. This does not seem quite right, as the former phrase (and “very plain,” also used as an equivalent in Sanofi), seems to be a higher standard than “fair expectation.” Near J then went on to say that the FCA decision in Pfizer v Apotex/ sildenafil (NOC) 2009 FCA 8, “intends that ‘fair expectation of success’ is the standard to be adopted by the Court. The Federal Court of Appeal, at para 44, described that ‘predictable’, and therefore obvious, solutions are equivalent to ‘solutions that provide “a fair expectation of success”’” [41]. But, citing Sanofi, the FCA also said that “the word ‘obvious’ in the phrase ‘obvious to try’ is to be given its primary meaning of ‘very plain’” [27]. Near J used both these terms in the somewhat confusing statement that “I thus conclude that it was self-evident or plain that there was a fair expectation that a sustained release formulation of quetiapine using HPMC would be successful” [43].

Perhaps the best way to reconcile these statements is to say that it must be “very plain” to try a candidate solution to the problem, but that solution need only have a “fair expectation of success.” So, if it was very plain that the combination of quietiapine and HPMC should be tried, and there was a fair expectation that it would succeed if tried, then the invention was obvious. This is consistent with Near J’s conclusion that “I conclude that it was more or less self-evident to try to obtain a sustained release formulation of quetiapine using HPMC, and that the person skilled in the art would have had a fair expectation of success” [66].

There is another ambiguity. Near J’s conclusion, just quoted, as well as other statements at [36], [41], suggest that there must be a fair expectation that the particular combination will succeed. But what if there are a handful of very plain possibilities, and a skilled person would not have a fair expectation that any particular one would succeed, yet there is a fair expectation of success with at least one amongst the group, and, in the end, one did succeed after only routine trials? This was essentially the case in ratiopharm v Pfizer Ltd. / amlodipine besylate. 2009 FC 711. Amlodipine had been delivered by the drug discovery unit to Pfizer’s development team as a maleate salt. The maleate had undesirable properties for commercialization, and it was well known that these properties, such as solubility, stability and stickiness, varied widely among the different salt forms. The properties of a particular salt could not be predicted. To identify a particular salt the patentee conducted a “salt screen” process in which a relatively small number of potential salts, chosen to represent a range of properties, were made and tested empirically for the properties of interest. Some of those salts proved unsuitable, but the best had properties that were acceptable for commercialization purposes, and a patent was obtained for one, the besylate salt, which was at issue. The salt screen process was routine, though the selection of the particular salts to be tested was a matter of judgment. The besylate was a standard pharmaceutical salt, and there were no particular problems encountered during the salt screen process. Thus, while the properties of the besylate salt could not be predicted in advance, so that it was not obvious that besylate would be suitable, it was obvious to try the screening process and there were no difficulties in conducting the process and arriving at the besylate salt. In summary, it was obvious to try and obvious to succeed, and therefore the claim was obvious, notwithstanding that the particular result was not obvious in advance. In that case, Hughes J held that the invention was obvious.

This issue was not raised in quetiapine, since there was a reasonable expectation that HPMC itself would succeed, but, following Hughes J’s decision in amlodipine besylate, the invention would likely have been obvious even if three or four of the most obvious gelling agents had been tried before finding one which worked (so long as the trials themselves were routine).

While my view is that Near J’s decision was clearly correct on the facts, it does raise a problem from an innovation perspective. Minor variants on an known drug may be obvious, but they may nonetheless be very useful, as in this case. We still want someone to come up with such variants, which may not be too difficult, but also to bring them through the regulatory process. The obviousness requirement which defeated the patent in this case may be justified on the basis that the patent system is designed to provide an incentive to invent, not an incentive to commercialize (though when there is an invention, both purposes are often served). But when there is no invention, where is the incentive to commercialize? The implicit presumption seems to be that there is a less of a free-rider problem in the post-invention commercialization process, but it is not clear that this is true in respect of pharmaceuticals. The data protection regime seems to offer a solution that is aimed exactly at protecting the investment in obtaining marketing authorization, where there is clearly a free-rider issue. But out of fear of evergreening, the data protection regime does not apply to variants of known drugs. This strikes me as a bit backwards. If you have a new drug, you can normally get patent protection, and so data protection is not needed. Data protection can fill a gap by providing a shorter term of protect for products that were relatively easy to invent, but which require an investment to get marketing approval. But that is exactly when data protection is denied.

Note that in this case the only points argued were obviousness and ambiguity, and Near J did not find it necessary to address the ambiguity argument in light of his conclusion regarding obviousness [67]. Note also that there were two parallel applications, relating to different dosages strengths, which gave rise to two sets of reasons 2013 FC 245/46, but the reasons themselves are identical.