Dependent Claims and Obviousness

Ciba Specialty Chemicals Water Treatments Ltd v SNF Inc 2017 FCA 225 Pelletier JA: Rennie JA, concurring reasons by Woods JA aff’g 2015 FC 997 Phelan J
            2,515,581

A post on the McCarty Tetrault snIP/ITs blog has pointed out an issue that I had missed in Ciba Specialty Chemicals Water Treatments Ltd v SNF. The authors note that the FCA held (my emphasis):

[97] The result is that Claim 1 is obvious and, since all other claims are directly or indirectly dependent on Claim 1, the ‘581 Patent is invalid.

They remark that

This holding is difficult to reconcile with prior Federal Court of Appeal jurisprudence not cited by the court. In Zero Spill Systems (Int’l) Inc. v. Heide [2015 FCA 115], the Federal Court of Appeal reversed the Federal Court on this very holding:

The Federal Court was obliged to consider the validity of the ’064 Patent claim-by-claim. Not doing so was a legal error. Notwithstanding that Claim 13 is actually an independent claim, the nature of dependent cascading claims is to narrow the claims upon which they depend: Purdue Pharma v. Pharmascience Inc., 2009 FC 726, 77 C.P.R. (4th) 262 at paragraph 10. The practical effect of this on anticipation or obviousness is that eventually a claim may be sufficiently narrow to escape these prior art-based attacks, even though the broader claims may be invalid. [94]

I suggest that the reason for the apparent conflict may be that in Ciba v SNF, according to Phelan J’s trial decision:

[26] As a result of the nature of the dependencies and the state of the art, if Claim 1 is invalid, then its dependent claims are likewise invalid. Dr Farrow, Ciba’s principal expert, was correct in his view that there was only one inventive concept that applied to all claims in the 581 Patent. This case rises or falls on Claim 1.

So, the FCA statement at [97] may only be saying that in this particular case none of the dependencies added anything potentially non-obvious. However, this is not entirely satisfactory. Phelan J analyzed obviousness in terms of the inventive concept, and since he concluded all the claims shared a single inventive concept, to hold that in this case all the claims rise or fell together (at least so far as obviousness is concerned), is not inconsistent with Zero Spill. But as discussed here, a key point of the FCA decision in Ciba v SNF was that it is preferable to “avoid[] the inventive concept altogether and pursu[e] the alternate course of construing the claim” [77], which is what the FCA did in undertaking its own obviousness analysis de novo. Since the FCA did not rely on the inventive concept at all in assessing obviousness, much less on there being a single inventive concept, it is not clear why the claims should stand or fall together.

I suspect that what happened is simply that on appeal the parties treated all the claims as standing or falling together, in light of Phelan J’s holding at trial, and consequently so did the FCA, even though the rationale had shifted. In any event, this particular point was tangential in Ciba v SNF, whereas in Zero Spill, as the snIP/ITs authors point out, it was the very issue on which the FCA reversed the court below. Moreover, the holding in Zero Spill is clearly correct as a matter of principle. All things considered, I think it may confidently be said that the FCA’s statement in [97] of Ciba v SNF, turned on the particular facts of the case, and the holding in Zero Spill remains the law.