Merck Sharp & Dohme B.V. v. Warner Chilcott Co. (Fed. Cir. 2017)

By Kevin E. Noonan --

The Supreme Court most recently revisited the proper standards for making an obviousness determination ten years ago, in KSR Int'l. Co. v. Teleflex. Inc.  While in some ways changing the obviousness standard, for example expanding the scope and relevance of analogous art, and clarifying other aspects, such as when being "obvious to try" can make a claim obvious, the Court's opinion was steadfast to established principles of patent law in other respects.  One of these latter principles is that the "pernicious effects of hindsight" should be avoided, for the distorting effect it can have on when an innovation is deserving of patent protection (and when it is not).

The District Court's failure to properly apply this principle of avoiding hindsight was the error that led the Federal Circuit on Friday to overturn a finding of obviousness in Merck Sharp & Dohme B.V. v. Warner Chilcott Co.  The invention related to an internally worn contraceptive device, sold by Merck under the brand name NuvaRing® as claimed in U.S. Patent No. 5,989,581, comprising an estrogen and a progesterone compound.  Contraception is effective when both compounds are delivered daily and consistently, but achieving this level of performance is the relevant challenge.  In the prior art, the two components (for Merck's device these are ethynyl estradiol (EE) and etonogestrel (ETO)) were combined in a single compartment which proved unsatisfactory so two compartment designs were used.  The invention claimed in the '581 patent provided both compounds in a single compartment wherein the polymer comprising the device contained supersaturated ETO.  Claims 4 and 11 were at issue, claim 4 being dependent on claim 1:

1.  A drug delivery system comprising at least one compartment which comprises
    a thermoplastic polymer core . . . said core comprising a mixture of a steroidal progestogenic compound and a steroidal estrogenic compound in a ratio by weight that allows a direct release of both said progestogenic compound and said estrogenic compound in physiologically required amounts,
    said progestogenic compound being initially dissolved in said polymer core material in a degree of supersaturation of 1 to about 6 times of the amount by weight necessary for obtaining saturation concentration of said progestogenic compound  in said polymer core material at 25° C,
    said estrogenic compound being dissolved in said polymer core material in a concentration lower than that of said progestogenic compound . . . .

4.  A drug delivery system according to claim 1, wherein the amount of progestogenic compound dissolved in the thermoplastic core material is 2 to 5 times the amount necessary for obtaining saturation concentration. (italics added)

And claim 11 being dependent on claim 5:

5.  A drug delivery system in a substantially ring shaped form and suitable for vaginal administration comprising at least one compartment which comprises
    a thermoplastic polymer core . . . said core comprising a mixture of a progestogenic steroidal compound and an estrogenic steroidal compound in a ratio by weight of 10 parts of the progestogenic compound to 1.5–5 parts of the estrogenic compound . . . .

11.  A drug delivery system according to claim 5, wherein the core material comprises 0.55 to 0.8% by weight of etonogestrel and 0.12 to 0.18% by weight of ethinyl estradiol. (italics added)

The important limitation recognized by the Federal Circuit is that the device comprises a single compartment that comprises both the progestogenic steroidal compound and the estrogenic steroidal compound (highlighted in italics above).

The District Court determined that these claims were obvious over the prior art disclosure of PCT Publication No. WO97/02015.  This PCT application was also directed to contraceptive vaginal rings but expressly disclosed a two-compartment structure.  The specification of the '015 PCT application disparaged one-compartment versions, stating:

These above-mentioned one-compartment rings have the disadvantage that, when loaded with more than one active substance, release patterns of these substances cannot be adjusted independently.  Such devices usually show suboptimum release patterns for the different substances, whereas it is generally preferred that all substances are released in a controlled rate and during a similar duration of time.  As a consequence the release ratio of the active substances undergoes a change after a period of time.

Nevertheless, the District Court found that the worker of ordinary skill would have been "motivated" to optimize the ring disclosed in the '015 PCT to include a second compartment wherein that compartment released the contraceptive-effective amounts of ETO and EE (presumably so that both compartments contained both compounds, thus distinguishing prior art comprising two components each containing only one of the compounds).  The '015 PCT disclosed drug release rates encompassing those of Merck's claimed device and thus "it would have been obvious for a person of skill to derive the claimed ratios of progestin and estrogen."

The Federal Circuit reversed, in an opinion by Judge Hughes, joined by Judges Linn and Dyk.  Taking the invalidated claims in order, the panel noted that the '015 PCT application disclosed a first compartment comprising EE alone and a second compartment a combination of EE and ETO.  The opinion disputes the District Court's finding that the '015 PCT application discloses a second compartment comprising 97% of the ring and includes ETO at a higher concentration than EE.  The '015 PCT application discloses ranges of these compounds ("between 30:1 and 1:30, but usually are between 15:1 and 1:1, and preferably are about 2:1.") and this can result in a ring wherein the second compartment comprises 97% of the ring.  However, this "hypothetical" ring is not disclosed and, according to the Federal Circuit is "outside of the usual or preferred range disclosed in PCT '015."  In addition, the panel opined that the District Court's interpretation would require the skilled worker to "pick a concentration of ETO from the high end of the disclosed range, but conversely select a concentration of EE from the low end of the range."  There is no disclosure in the '015 PCT application (the only art the District Court relied upon) that would suggest to the skilled worker to choose these concentrations.  The panel found that "the only way to arrive at the hypothetical ring is by using the '581 patent as a roadmap to piece together various elements of PCT '015."  And doing that, according to the Federal Circuit, is hindsight.

This conclusion is supported for the panel by the disparagement of one-compartment embodiments expressly contained in the '015 PCT application's specification.  The opinion distinguishes the instructions from the Supreme Court's KSR opinion that the skilled work may sometimes "have 'good reason to pursue the known options' based on 'design need or market pressure.'"  But those instances involve ones where the skilled worker pursues "identified, predictable solutions," whereas here the skilled worker would have had to rely on putatively inoperative embodiments, which the very prior art the District Court relied upon provided devices having "sub-optimal release patterns."

With regard to claim 11, which does not require effective amounts of both compounds to be released from one compartment, the amounts of the compounds are specified in the claim.  Warner-Chilcott argued persuasively to the District Court that the skilled worker would have been able to calculate these amounts based on the release rates disclosed in the '015 PCT application.  The Federal Circuit opinion points out that these release rates were disclosed in the '015 PCT application as being determined using a two-compartment device.  The opinion states that its reading of the '015 PCT application is that these disclosed release rates can be achieved because they are contained in two compartments.  Thus, for the panel the skilled worker would not have expected to be able to produce these release rates from a one-compartment device.

This decision is nonprecedential, but provides a good example of how the Federal Circuit understands the proper avoidance of hindsight in considering whether a claim is obvious.

Merck Sharp & Dohme B.V. v. Warner Chilcott Co. (Fed. Cir. 2017)
Nonprecedential disposition
Panel: Circuit Judges Dyk, Linn, and Hughes
Opinion by Circuit Judge Hughes

Comments

[maxdrei]

I do not see any reliance on the notion of "teaching away" in the Report above, or in the text of the Decision. But it's there though, isn't it? See the para that begins "The district court's reliance onm hindsight is further underscored by....."

I mention this because of the discussion of what "teaching away" means, on another well-known patent law blog recently.

Where I write from, Europe, we don't do "teaching away". We do recognise, however, a "technical prejudice" against that which is claimed can evidence non-obviousness.

There are tons of reasons why an individual patent application might "teach away" from something, and few of them prove that any "technical prejudice" exists. So I'm deeply sceptical that one patent publication "teaching away" is a solid enough reason for deciding that the thing claimed is not obvious.

[Kevin E Noonan]

Thanks, Max - while "teaching away" is one way of making the point, the better characterization is disparagement (a higher order form of teaching away). The distinction is that while a reference can say that one thing is better than another, or that a combination is unlikely to work (teaching away), here the portion of the '015 PCT quoted in the post goes further (in my opinion). A judgment call.

Thanks for the comment

[MaxDrei]

Kevin thanks. The notion of "disparagement" I find useful to discuss. Call me cynical, but it leaves me even less of a fan.

I want (if it is possible) to patent ALL practical ways of implementing a concept, the concept as such being unpatentable. Say, for example, that I can implement with an "odd number" implementation or with an "even number" implementation. I write two patent applications, one extolling the odd number solution. I deliberately include all the disparagement of even numbers that I can dream up. My other patent application extols the even number solution to the technical problem, and persuasively disparages the odd number solution.

Each patent application helps the other to issue. Bingo!

One can implement PTCA (known as such) with a balloon that is non-compliant, or with a compliant balloon. Perhaps then two patent applications are called for, one of which disparages compliant balloons, the other non-compliant balloons. Who amongst my competitors does not infringe one or other of my patent Tag Team?

Or a medication-dispensing ring. Two options. One chamber for the medication. Or two. Use disparagement to cover both options?

These are thoughts that leave me leary of building a case of non-obviousness around statements in a published patent document that can be fairly characterized as disparaging.

As you say, a judgment call. Is it plausible that the notional PHOSITA "disparage" the solution to anything like the same extent?

[Kevin E Noonan]

Well, Max, disparagement isn't just name calling - there needs to be some disadvantageous property (inoperability is usually the basis) for the alternative. It is also used to greatest effect as here (and in your example) when there is a dichotomy - one or two, black or white, odd or even - and it has the effect (as here) of precluding the alternative. So that if the accused infringer used two chambers the patentee would be estopped from raising a successful infringement action (which is why it is a tricky strategem).

As for your example, in the US the patentee suing on either patent would have the statements in the other patent raised in the litigation and might be guilty of inequitable conduct.

Thanks for the conversation