Sanofi-Aventis U.S. LLC v. Sandoz, Inc. (Fed. Cir. 2009)

More Claim Construction Confusion from the Federal Circuit

    By Kevin E. Noonan --

The Federal Circuit demonstrated once again the quagmire that can exist when lower courts attempt to apply the CAFC's rather quixotic jurisprudence on claim construction.

The case, Sanofi-Aventis U.S. LLC v. Sandoz, Inc.⁽¹⁾, arose as the result of ANDA filings by several Defendants for oxaliplatin, the active ingredient in Sanofi's Eloxatin® colorectal cancer drug.  Two claims of the patent at issue, U.S. Patent No. 5,338,874, were asserted against the Defendants:

1.  Optically-pure cis-oxalato(trans-l-1,2-cyclohexanediamine) Pt(II) having the general formula of Formula (1).

(Dependent claim 2 merely added a melting point limitation and was not further at issue in the case.)  The question decided by the Federal Circuit on appeal was whether the claim term "optically pure" included an affirmative limitation that "optically pure oxaliplatin" had been "resolved by means of the HPLC method described" in the specification of the '874 patent.

The District Court had decided that this term did include the limitation.  While admitting that "[t]here was no dispute that nothing on the face of the claims of the '874 patent limit[ed] the claims to 'optically pure oxaliplatin that is produced through the use of HPLC," the District Court identified statements in the prosecution history and disclosure in the specification that supported its interpretation.  Specifically, the District Court was persuaded that Sanofi had distinguished an asserted prior art reference (Kidani et al.) that purported to disclose "a single isomer [of oxaliplatin] useful as an antitumor agent" by arguing that the optical purity of the oxaliplatin in the reference was only 90%.  According to the District Court, Sanofi stated during prosecution that "[o]nly after HPLC resolution (in accordance with the teachings of the present application) was optical purity obtained."  In addition, the District Court pointed to portions of the specification that compared the optical purity of oxaliplatin prepared according to prior art methods contained in the asserted reference with the optical purity of Sanofi's molecule after resolution by HPLC.  To the District Court, these distinctions were sufficient to require that the claim term "optically pure oxaliplatin" describe only oxaliplatin resolved to optical purity by HPLC.  Accordingly, since the Defendants did not use HPLC to resolve oxaliplatin to optical purity, the District Court granted summary judgment of non-infringement.

The Federal Circuit disagreed with the lower court's claim construction.  In a unanimous (albeit nonprecedential) opinion by Judge Moore, joined by Judges Linn and Prost, the Court found that the patentee had not disclaimed all embodiments of optically pure oxaliplatin except those resolved by HPLC.  The Court opined as follows:

We have repeatedly warned of "the danger of reading limitations from the specification into the claim."  See, e.g., Phillips, 415 F.3d at 1323.  "Absent a clear disavowal or contrary definition in the specification or the prosecution history, the patentee is entitled to the full scope of its claim language."  Home Diagnostics, Inc. v. Lifescan, Inc., 381 F.3d 1352, 1358 (Fed. Cir. 2004).  To narrow the plain language of a claim, a disclaimer must be clear and unmistakable.  Cordis Corp. v. Boston Scientific Corp., 516 F.3d 1319, 1329 (Fed. Cir. 2009).  We see no such disclaimer in the specification or prosecution history of the '874 patent.

The Court also disagreed that claim 1 of the '874 patent was a product-by-process claim:

The district court relied on Andersen Corp. v. Fiber Composites, L.L.C., 474 F.3d 1361 (Fed. Cir. 2007), when construing claim 1 as a product-by-process claim.  In Andersen, this court held that claims to composite structures included a pelletizing process limitation where the patentee relied on that process both to define the invention and to distinguish the prior art.  . . .  We determined that the specification attributed the claimed physical properties to the process and that the specification indicated that the pelletizing step was a requirement, not a preference, of the invention.  . . .  We further determined that the patentee had clearly disavowed other processes during prosecution.  . . . [emphasis added]

By contrast, here, the patent specification and prosecution history focus on the property of the composition (optical purity) and not the process used to obtain that property.  . . .  The specification never asserts that HPLC is required to obtain optically pure oxaliplatin.  It characterizes HPLC as an "illustrative method" and a "representative process" by which the claimed compound "may be prepared".  . . .  Moreover, the specification does not define the property (optical purity) by reference to the process of purification by HPLC.  Thus nothing in the specification limits the invention to optically pure oxaliplatin purified using HPLC.

The prosecution history also illustrates that it is the optical purity of oxaliplatin that distinguished it from the prior art, not the process used to obtain that purity.  The Examiner rejected the initially filed claims to oxaliplatin "of optically high purity" as anticipated or rendered obvious by Kidani.   The Examiner stated that Kidani disclosed "a single isomer [oxaliplatin] useful as an antitumor agent.  Note that since the single isomer complex was prepared, the optical purity of such material is very high or almost pure isomer."  In response, the applicant . . . explained that it had repeated Kidani's process "using identical reactant materials and the subsequent testing thereof. . . . The resultant material was tested and found to be 90% [oxaliplatin] not optically pure . . . ."  Sanofi explained that "[o]nly after HPLC resolution (in accordance with the teachings of the present application) was optical purity obtained."  Sanofi further asserted that the products prepared using Kidani's method "do not have the presently claimed optical purity."  Therefore, Sanofi argued that the claimed oxaliplatin "having high optical purity[] is not found or taught in the prior art either by inherency or by being obvious thereover."  Following a telephone interview, Sanofi agreed to amend the claims to "optically pure" oxaliplatin, rather than oxaliplatin "of high optical purity."  The Examiner entered the amendment and allowed the claims, stating that "[t]he Examiner agrees with applicants that Kidani et al. does not teach[] the cis-oxalato(trans-l-1,2-cyclohexanediamine)Pt(II) as an optically pure isomer.  It is clear from Kidani et al. that also other isomers can be in the final product."  Thus, Sanofi argued that the defining feature of the claimed oxaliplatin was its optical purity, not the HPLC process.  Nothing in the prosecution history amounts to a clear and unmistakable disclaimer of optically pure oxaliplatin prepared using other (non-HPLC) processes.  [emphasis added].

The Federal Circuit vacated the District Court judgment of non-infringement and remanded the case to the lower court.

The panel's opinion illustrates how difficult it is for a district court to implement supposedly straightforward rubrics such as those enunciated by the en banc court in Phillips v. AWH Corp.  The District Court's reasoning was based on understanding the plain meaning of the term "optically pure," a phrase that has no plain meaning outside its specialized meaning in the chemical arts.  The District Court looked to the specification and the prosecution history to identify how the patentee defined the term, and in the absence of an express definition used the examples in the specification and the arguments and amendments made during prosecution to construe the claim term "optically pure" to require that optical purity must be achieved using HPLC.  The Federal Circuit disagreed and substituted its judgment for the lower court's, a commonplace occurrence.

The outcome brings to mind the dissenting opinion of District Court Judge Vaughn R. Walker, Chief Judge, U.S. District Court for the Northern District of California, sitting by designation, in Medegen MMS, Inc. v. ICU Medical Inc. (Fed, Cir. 2008) (see "Claim Construction at the Federal Circuit: A District Court Judge's View").  In that case, Judge Walker asserted that claim construction presented "a conundrum" to district courts.  This conundrum arose from these "twin axioms" of claim construction:  "[o]n the one hand, claims 'must be read in view of the specification'," citing Markman, while "[o]n the other hand, it is improper to read a limitation from the specification into the claims," citing Arlington Indus. Inc. v. Bridgeport Fittings, Inc.  The problem with these axioms, according to Judge Walker, is that "the axioms themselves seldom provide an answer, but instead merely frame the question to be resolved."  While recognizing the sound policy reasons for a patent drafter to be required to include "all claim limitations in the claims of a patent," he also admonished that "patents must be read as well as written."  Citing Phillips, he asserted that this analysis must be performed with due consideration of a term's "ordinary and customary meaning" informed by the "context of the entire patent, including the specification."

Judge Walker's dissent raised the essential paradox of the Federal Circuit's practice under Cybor Corp. v. FAS Technologies, Inc. of giving no deference to claim construction by district courts:  "[a]s a jurist more accustomed to working on the front lines of patent litigation than reviewing decisions from above, it is my experience that claim construction -- determining how one of ordinary skill in the art would understand the patent at the time of invention -- often requires making fact-like determinations not well suited to appellate review."  The "uniformity and predictability" desired by the CAFC as a general, guiding principle is not always best served in the individual determinations necessary to construe claim terms.  In this he cited with approval Judge Lourie's point in Phillips, that while there is (and perhaps cannot be) formal deference by the CAFC to a district court's claim construction, it is better policy for the Federal Circuit to affirm "in the absence of a strong conviction of error."

Current academic research suggests that Federal Circuit affirmance rates of district court claim construction has "improved" to from 60-70%.  However, each Federal Circuit panel addresses, and resolves, specific claim construction issues in any particular case in its own way, introducing the kind of uncertainty that the Federal Circuit was created to curtail.  The inconsistencies resulting from the Federal Circuit's "no deference" standard of review can be expected to continue to produce the type of results decried by Judge Walker in Medegen and, arguably, that were obtained in Sanofi-Aventis v. Sandoz.

⁽¹⁾ Additional defendants included Teva Parenteral Medicines, Inc., Teva Pharmaceuticals USA, Inc., and Pharmachemie BV; Mayne Pharma Ltd., Mayne Pharma (USA) Inc,, Hospira Australia Pty Ltd., and Hospira, Inc.; W.C. Heraeus GMBH; APP Pharmaceuticals, Inc. and Abraxis Bioscience, Inc.; Actavis Totowa LLC, Actavis, Inc. and Actavis Group HF; Fresenius Kabi Oncology PLC and Fresenius Kabi Pharma Ltd.; Sun Pharmaceutical Industries Ltd. and Caraco Pharmaceutical Laboratories, Ltd.; Ebewe Pharma GES.M.B.H. NFG HG; and Mustafa Nevzat Ilac Sanayii A.S., Par Pharmaceutical Companies, Inc. and Par Pharmaceutical, Inc.

Sanofi-Aventis U.S. LLC v. Sandoz, Inc. (Fed. Cir. 2009)
Nonprecedential disposition
Panel: Circuit Judges Linn, Prost, and Moore
Opinion by Circuit Judge Moore

Comments

[Dan Feigelson]

Kevin, I agree that it's sometimes difficult for district courts to distinguish between permissibly relying on the specification to construe claims, on the one hand, and impermissibly reading limitations from the specification into the claims on the other hand. But I don't think case illustrates that difficult - this one isn't even a close call. The claim recites an optically pure compound. The specification inherently defines optically pure as being free from the other enantiomer (100% enatiomeric excess), so there's no indefiniteness issue. And the specification teaches at least one method to make the optically pure material, viz. via HPLC, so there's no written description or enablement issue. That should have been the end of the discussion. There may or may not be other methods to make the claimed material, but that's irrelevant, because anyone who makes or possess the claimed material infringes. Note also that the spec says a separate application was filed directed to the purification method per se (although I didn't investigate the fate of that application), which would seem to auger for a process-free claim construction in the patent at issue.

[EG]

Kevin,

"Quixotic jurisprudence on claim construction" eh? We are waxing "lyrical" today aren't we?

I do agree with you that the way the Federal Circuit handles claim construction can be very frustrating for district court judges. And Judge Newman has written some fairly "volcanic" dissents about how unrealistic it is to treat claim construction so much as a question of law, which is what Cybor commands. That, like this case, the Federal Circuit so often parses through the specification and file history suggests otherwise.

[Kevin E. Noonan]

Dear Dan:

I don't dispute the CAFC's determination, just that claim construction is a place where reasonable people can differ. One way of looking at the spec and prosecution history is the way you and the court did. Another is to interpret the evidence as meaning that HPLC was the only way (or asserted to be the only way) to achieve this degree of optical purity. I don't know the answer, and I don't know if the district court considered any evidence on this point.

But my point is that these types of questions are best considered at trial, where testimony can be elicited on this topic. Clearly, if HPLC was but one tool in the chemist's toolbox for making optically pure oxaliplatin, then the district court made an error in putting so much weight on the fact that the patentee chose to use HPLC to go from 90% pure (the prior art) to 100% pure. But we don't know if that is the case; here, the issue was framed as one of pure claim construction.

Having claim construction at the district court be but a way-station to Federal Circuit review is a bad idea on many fronts. It is excessively costly, it delays final resolutions, it promotes premature summary judgment decisions, and it puts in the appellate court's hands questions that have sufficient factual predicates that some assistance from the district court might prove helpful in many cases. But why should a district court put in the time and effort to "get it right," if it knows any decision it makes will be second-guessed and likely overturned?

Finally, the current system makes patent claim meaning a crap shoot of which panel of the court you happen to get. This wasn't the increase in certainty that Congress wanted when it created the court. (Ironically, the Federal Circuit's animus against the doctrine of equivalents stems from the court's reluctance to have patentees introduce a similar level of uncertainty to patent claim "construction.")

I have no opinion about whether the district court or the Federal Circuit "got it right" - there are arguments for both positions. But I think the discussion about the process is important, so thanks for contributing.

[Kevin E. Noonan]

Dear EG:

What do you have against lyricism once in a while?

[EG]

Dear Kevin,

None, I was being very "tongue in cheek" on that! And I too waxed "lyrical" on the Cardiac Pacemarkers case. See IPWatchdog.

[MaxDrei]

Just a thought but is the problem not that the Federal Circuit regards itself as somehow better qualified than the District Court judges to "do" claim construction properly, and that despite the factual background to claim construction, established by witness evidence at trial.

To me, from England, it seems very simple. Claim construction is a matter of law, so you need an appeal instance that is competent to lay down clear law, District Court judges who are competent to follow that law, and (to complete the virtual circle) CAFC judges who defer to competence, and reverse only when they feel they have no choice.

Otherwise, with rules that contradict each other, it's nothing but a downward spiral.

[Kevin E. Noonan]

Dear Max:

No doubt, but the practice of having district court judges "sitting by designation" may help.

I think rather than a downward spiral it's more of a sine wave (or a roller coaster), but in any event the present practice doesn't promote certainty.

Thanks for the comment.