Tenofovir is a drug used in the treatment of HIV/AIDS. The ‘619 patent claims tenofovir disoproxil, a prodrug of tenofovir, while the ‘059 patent claims the fumarate salt of tenofovir disoproxil . The main issue in respect of both patents was invalidity for obviousness, and in particular whether the invention was “obvious to try.” Barnes J held the ‘619 patent to be valid, and the ‘059 patent to be invalid. In my view, his application of the obvious to try analysis was correct, but the case reveals a tension in the obvious to try test which has emerged particularly since the FCA decision in Sanofi-Aventis v Apotex Inc / clopidogrel 2013 FCA 186 (blogged here).
In Sanofi 2008 SCC 61 the SCC held that an “obvious to try” test may be warranted as part of the obviousness analysis, and one factor to be considered is “Is it more or less self-evident that what is being tried ought to work?” . This phrase contains an important ambiguity. Does “what is being tried” refer to the particular claimed invention? That is, to establish obviousness it is necessary that a posita could have predicted in advance of experimentation that the claimed invention would solve the problem at hand? Or does “what is being tried” refer simply to a solution to the problem? That is, it is more or less self-evident that a posita would be able to solve the problem at hand, even though it may not have been obvious prior to experimentation that the particular claimed invention would be the solution? For example, if the problem is to resolve a racemate into enantiomers, must it be predictable in advance that the claimed method would work, or it is sufficient if the claimed invention is one of three standard methods, and it is plain that one of them would work, even if not predictable which method in particular would be successful? The first, narrower, approach is more favourable to the patentee, while the broader approach is more likely to result in the invention being found obvious. (I should repeat that this question, whichever is correct, does not exhaust the obvious to try analysis; it is only one factor to be considered.)
In my view, the second, broader approach is correct. The obvious-to-try test is particularly applicable “[i]n areas of endeavour where advances are often won by experimentation” [Sanofi 68]. The whole point of the fact that advances are “won by experimentation” is that a priori prediction on the basis of theory is impossible. The theory may suggest certain avenues to explore, but success remains uncertain until the experiment is made. So, in the seminal decision of the EWCA in Johns-Manville’s Patent  RPC 479 the claim was to a flocculating agent (which causes separation of suspended solids) for use in production of asbestos cement. It was well known that a flocculating agent that would not weaken the cement would be desirable, but none had previously been found. The flocculating agent in question had recently been developed and used with success in other industries, so it was held to be obvious to try it in making asbestos cement. There is no suggestion in that decision that it could have been predicted with any particular expectation of success that the flocculating agent in question would work, apparently because the question of whether the agent would weaken the cement was a question that could only be answered by experiment. The patentee’s argument was precisely that the invention was
only obvious if success could be predicted in advance (485), and this was the test that was rejected by the Court of Appeal (495): see my case comment "Restricted application of `obvious to try'" (2012) 7 JIPLP 304; doi: 10.1093/jiplp/jps030.
Nonetheless, in its Clopidogrel 2013 FCA 186 the majority of the FCA adopted the more restrictive approach to the obvious to analysis, saying (original emphasis):
 The focus of the obviousness analysis in Plavix [Sanofi SCC] was not the difficulty in seperating the racemates covered by the `875 genus Patent - which included PCR 4099 - but the unknown properties of the resulting enantiomers. :
The method to obtain the invention of the `777 patent were common general knowledge. It can be assumed that there was a motive to find a non-toxic efficacious product to inhibit platelet aggregation in the blood. However, it was not self-evident from the `875 patent or common general knowledge what the properties would be and therefore that what was being tried ought to work.
With respect, the SCC Sanofi decision is not entirely clear on this point. So, immediately after asking, “Is it more or less self-evident that what is being tried ought to work?” the Court elaborated by asking “Are there a finite number of identified predictable solutions known to persons skilled in the art?” . This suggests the broader interpretation: if there are a small number of solutions, one of which is likely to solve the problem, this is enough. See Gauthier J’s concurrence in Clopidogrel, suggesting a broader interpretation of Sanofi. But there are other passages in the SCC decision, relied on by the majority of the FCA in Clopidogrel which can be read to support the narrower position, as I discuss in my post on the Clopidogrel decision.
On the facts of the case at hand, tenofovir was known to be an effective antiviral agent useful to treat HIV/AIDS , but it had poor bioavailability . It was well known that one way the bioavailability of an API can be improved is the development of a prodrug, which is a compound which is itself inactive, but has better bioavailability, and which is metabolized into the active drug once the delivery barrier are crossed . Tenofovir disoproxil is a prodrug of tenofovir which uses the carbonate promoiety (the inactive half) disoproxil .
In setting out the law, Barnes J stated that “[t]he strength of the ability to predict success is the lynchpin to an obvious to try analysis” and he quoted extensively from the FCA Clopidogrel decision, including , in setting out the law . This suggests that he was adopting the narrower question, which would ask whether it was self-evident that disoproxil would work to create an effective prodrug. It would be very difficult to establish obviousness on this basis. However, while Barnes J did hold that the invention of ‘619 patent was not obvious, he did not state the question in this precise form, nor did he rely on the fact that the properties of tenofovir disoproxil were not self-evident. Instead, he emphasized the uncertainty and difficulty involved in prodrug development generally. For example, he accepted the view expressed in a review article that “choosing a suitable bioreversible protective group for phosphates, phosphonates [a class including tenofovir] and phosphinates presents a major challenge,” and “further and more innovative prodrug research” is required . Barnes J’s extensive analysis of the expert evidence, and Gilead’s actual course of conduct, all supported the same conclusion. Consequently, it seems clear to me that if Barnes J had asked the broader question – is it self-evident that a prodrug could be created? – he would also have answered in the negative. So, while Barnes J formally accepted the narrow interpretation in Clopidogrel as the law, his decision with respect to the ‘619 patent is consistent with the broader approach to the obvious to try test.
Barnes J then turned to the ‘059 patent, which claims the fumarate salt of tenofovir disoproxil . The question is whether the choice of the fumarate salt was obvious . Barnes J initially posed the question as follows:
 The validity of the patent turns on whether using fumaric acid as a salt former with bis(POM)PMPA was “obvious to try” and whether it was more or less self-evident that a suitable pharmaceutical salt would result.
This suggests the narrower approach: the question is whether it could be predicted that fumaric acid in particular would result in a suitable salt. However, after determining that “that fumaric salt would be, at least, one of several obvious salt formers the person of skill would try,” , Barnes J said:
 This does not end the inventiveness inquiry. I must also assess whether it was more or less self-evident that the inventors would successfully arrive at a pharmaceutically acceptable salt of tenofovir disoproxil.
This is a clear statement of the broader formulation of the obvious to try test; the question is not whether it was self-evident that the fumarate salt in particular would work, but whether it was self-evident that some salt would work. On the facts, Barnes J concluded that
 Although a person of skill may not have predicted with a high degree of certainty that fumaric acid could be used to produce an acceptable salt formulation for tenofovir disoproxil, there would still be an expectation that, with routine screening of a handful of acidic salt formers, one or more acceptable compounds would emerge.
Here, he expressly rejects the proposition that it must be predictable that fumaric acid itself would work. He nonetheless held that the ‘059 patent was invalid for obviousness. This is really only consistent with the broader approach to the obvious to try test, which, in my view, is the correct approach. Barnes J’s conclusion on this point is consistent with Hughes J’s Amlodipine decision 2009 FC 711, which arrived at the same conclusion on similar facts: see my discussion here.
In conclusion, while the obvious to try test is now well established in Canadian law, significant uncertainty remains. It now seems to be established that a salt form developed by a routine salt screen will be obvious, even if the particular salt which would be successful could not be predicted without experimentation. (This is not to say that all salt forms are obvious, as in some cases the screen or the choice of the candidate salt may not be routine.) In my view, these salt screen cases are paradigmatic of the proper application of the obvious to try test. The question is whether this will be seen as a special rule related to salts, or whether the broader principle underpinning the salt screen cases will get leveraged into a more general context.
Note: Bristol-Myers Squibb v Teva Canada Ltd 2013 FC 1271 [ATRIPLA] and Gilead Science Inc v Teva Canada Ltd 2013 FC 1272 [VIREAD] are companion cases arising because Gilead markets two drugs containing tenofovir disoproxil, and BMS markets a third, and three separate applications were commenced in respect of these drugs. The issues were identical, and it was agreed that the decision reported in 2013 FC 1270 would apply to these applications as well, which were therefore similarly decided -.