Teva Canada Limited v Novartis Pharmaceuticals Canada Inc, 2016 FCA 230
It is often good practice for a patent filer to load up a patent specification with many different implementations of their invention, even if each implementation has its own slightly different advantages, in hopes to obtain the broadest patent protection possible in a single patent. This case serves as further assurance that this practice is a sound one, even when considering the feared “promise doctrine” – if the patent drafting is handled with care.
In this case, Teva Canada Limited (“Teva”) failed to show that the Canadian patent on EXJADE, owned by Novartis Pharmaceuticals Canada Inc. (“Novartis”), is invalid for failing to deliver on what it argued was an overarching promise of utility. Teva argued that there is an overarching promise in the EXJADE patent that the “Formula I” and the “Formula II” compounds, as classified in the EXJADE patent, each treat certain diseases in humans.  Due to some excellent patent drafting, the Federal Court of Appeal (“FCA”), however, found that the EXJADE patent was drafted so as to make an important distinction between the utilities of the two sets of compounds, [11-12] and thereby held the Formula II claims to a lesser promise, and dismissed Teva’s allegations of inutility. 
Patent Drafting re Formula I and Formula II Compounds: Different Utilities & Different Promises
It was agreed by the parties that the Formula I compounds contain some compounds that had been previously made.  The novelty of these known compounds, as described in the EXJADE patent, was unequivocally in their use for the treatment of diseases which cause excess metal in the human body.  There was therefore a promise that the Formula I compounds be useful in the treatment of such diseases.
The Formula II compounds, on the other hand, are a subset of Formula I compounds, but are said to be themselves novel, having never previously been made. [14-5] Teva argued that the treatment promise should be applied to the claims for the Formula II compounds just as the promise was applied to the claims for the Formula I compounds. 
Teva’s argument was not supported, however, by the fact that all of the statements of use in the disclosure relate only to the Formula I compounds. The use of the Formula II compounds was not discussed in the EXJADE patent, other than in the following passage: 
It has now been found that certain substituted 3,5-diphenyl-1,2,4-triazoles have valuable pharmacological properties when used in the treatment of disorders which cause an excess of metal in the human or animal body or are caused by it, primarily a marked binding of trivalent metal ions, in particular those of iron. 
The phrase “certain substituted 3,5-diphenyl-1,2,4-triazoles” was interpreted to be an umbrella term referring to both Formula I and Formula II compounds.  Although this phrase appeared to make a promise and used the umbrella term, the promise was interpreted to apply only to Formula I compounds.
Teva argued that since every Formula II compound is within the superset of the known Formula I compounds, the novel Formula II compounds should be subject to the same promise made of the treatment of humans via the Formula I compounds.  The FCA rejected this argument on the principle that different claims can have different utilities, even for the same compound.  Furthermore, given the manner in which the patent disclosure and claims differentiated between Formula I and Formula II compounds, [24-25, 28] the FCA determined that the person skilled in the art would be able to reasonably read the patent as excluding Teva’s elevated promise of utility, [26-27] and therefore sided with Novartis by refusing to find that the Formula II compounds promised the same utility.
The promise of the Formula II compounds, then, was merely their capacity to bind iron.  Considering this lower threshold, the FCA confirmed the FC’s finding that the stated utility had been demonstrated or soundly predicted and that Teva’s allegation of inutility was unjustified. 
A patent is supposed to be granted for a single invention only, but the EXJADE patent describes at least two different sets of molecules with two different uses. This may to some extent explain the point of view behind Teva’s assertion that there should be only one overarching promise of the patent. But a patent cannot be invalidated solely because it discloses more than one invention (and indeed this was not argued), and it has been well accepted that each claim in a patent is construed independently. There therefore appears to be no contradiction in a patent having multiple claims each with different utilities and, sometimes, with different promises.
Practically speaking, this means that, even from the perspective of protecting your patent from the “promise doctrine”, patent filers can continue to load up a patent specification with many different implementations of an invention, each with its own advantages, so long as the patent description is properly drafted so as to maintain a distinction between each of the implementations and its advantages.