Last week in Boston Scientific v. Cordis, the Court of Appeals for the Federal Circuit declared invalid several claims of a Boston Scientific patent covering drug-eluting stents. While the conclusion of obviousness might not be surprising, the analytical path leading to the conclusion is very interesting and is worthy of attention.
Overturning a jury verdict holding the claims valid and infringed, the court determined that the claims were obvious in light of a single prior art patent. The reference taught all limitations of the claims, but did so across two separate embodiments. To arrive at the claimed invention, the court reasoned, a skilled artisan need only combine the two embodiments, which were illustrated in side-by-side figures in the patent. According to the court, making this combination “does not require a leap of inventiveness.”
So far, this seems fairly straightforward. But here’s the interesting part – the court did not conduct any sort of analysis under Graham v. John Deere to support its conclusion of obviousness. The opinion has no discussion of the Graham factors and, indeed, doesn’t even cite the case. Instead, the court points to the “predictable variation” language of KSR v. Teleflex and indicates that “one of ordinary skill in the art would have been motivated to combine the [separate embodiments of the reference] to arrive at [the claimed invention].”
Is it acceptable to skip the Graham analysis under these circumstances? The “predictable variation” language of KSR – relied upon here as support for the conclusion of obviousness – seems to cry out for a Graham analysis:
“[i]f a person of ordinary skill can implement a predictable variation, §103 likely bars its patentability.” (emphasis added)
“Likely” seems to leave room for the possibility that a claim to a predictable variation is, in fact, nonobvious. The Graham framework likely provides an appropriate structure for evaluating the issue – secondary considerations seem particularly relevant here. The court, nevertheless, skipped the analysis altogether, at least in the written opinion.
This seems to beg the question of whether a summary conclusion of obviousness is acceptable in certain situations, or whether a full Graham analysis is always necessary. Is this the summary conclusion of obviousness that KSR warned against?
There’s another interesting wrinkle to the Boston Scientific v. Cordis opinion – While the court failed to mention or discuss the Graham factors, it did note – in two separate instances – that the skilled artisan would have been motivated to combine the separate embodiments of the prior art patent to arrive at the claimed invention. In the absence of a Graham analysis, the mentions of a ‘motivation to combine’ appear to be an important factor on which the court relies for its conclusion of obviousness. Indeed, in the absence of any discussion of the Graham factors, it seems to come dangerously close to litmus-test treatment of the issue.
Does this hint at a clarification of the role of the TSM test in an obviousness analysis in the post-KSR world? If, in fact, the ‘motivation to combine’ was used in litmus test fashion, is this acceptable following KSR? All of these questions surrounding Boston Scientific v. Cordis are likely limited to the realm of obviousness issues based on combinations of embodiments from a single reference, which is presumed to be a relatively small universe.
Read the PTP review and analysis of Boston Scientific v. Cordis here.
