Unigene Laboratories, Inc. v. Apotex, Inc. (Fed. Cir. 2011)

By Kevin E. Noonan --

The Federal Circuit today issued its ruling in Unigene Labs., Inc. v. Apotex, Inc., and the decision illustrates the extent of the effects of the Supreme Court's decision in KSR Int'l Co. v. Teleflex Inc. on the Federal Circuit's obviousness jurisprudence.  In addition, the decision provides a contrast between how the Court views chemical obviousness for pharmaceutical formulations today, and the views contained in the Court's opinion in Pfizer, Inc. v. Apotex, Inc.

The case involved litigation over Unigene's drug Fortical®, a nasal formulation of salmon calcitonin for treating post-menopausal osteoporosis.  This drug is an alternative formulation of another salmon calcitonin-based drug, Miacalcin®, sold by Novartis.  The patent-in-suit was Unigene's U.S. Patent No. RE40,812, a reissue of U.S. Patent No. 6,440,392, and specifically asserted claim 19:

A liquid pharmaceutical composition for nasal administration comprising about 2,200 MRC units of salmon calcitonin, about 20 mM citric acid, about 0.2% phenylethyl alcohol, about 0.5% benzyl alcohol, and about 0.1% polyoxyethylene(2) sorbitan monooleate.

The District Court noted that calcitonin was known in the art to be difficult to administer, since it was known to be readily degraded, unstable, and poorly absorbed.  Both Miacalcin® and Fortical® contain 2200 IU salmon calcitonin, but the formulations differ:  the Novartis formulation contains sodium chloride, nitrogen, hydrochloric acid, water, and benzalkonium chloride (to promote absorption), while Unigene's formulation contains 20mM citric acid, polyoxyethylene(2) sorbitan monooleate (also termed "polysorbate 80"), phenylethyl alcohol, and benzyl alcohol.  Unigene received FDA approval of Fortical® under an NDA pursuant to 21 U.S.C. § 355(b)(2) using Miacalcin® as "reference drug."

Litigation ensued after Apotex filed an ANDA containing a Paragraph IV certification.  In its certification, Apotex asserted noninfringement, invalidity, and unenforceability for inequitable conduct.  The District Court found (on summary judgment) that the claims were nonobvious over "forty four prior art references" and that Apotex had not adduced sufficient evidence of an intent to deceive to support the inequitable conduct allegation; in this regard, the Court also refused to breach the attorney-client privilege under the crime-fraud exception.  The key for the District Court's determination of nonobviousness was that the art did not disclose using 20 mM citric acid in the formulation "to achieve 'both shelf stability and enhanced bioavailability' in a nasal salmon calcitonin formulation."  With regard to the other affirmative defenses, the District Court "held that all of Apotex's defenses and counterclaims . . . had been conceded, waived, barred, abandoned, or improperly raised."

The Federal Circuit affirmed in an opinion written by Chief Judge Rader and joined by Judges Moore and O'Malley.  In the opinion, the Court recited several rubrics from KSR that showed the nuances developed by the Federal Circuit in its application of the Supreme Court precedent.  These include:

• "Obviousness requires more than a mere showing that the prior art includes separate references covering each separate limitation in a claim under examination.  KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 418 (2007)."

• "Rather, obviousness requires the additional showing that a person of ordinary skill at the time of the invention would have selected and combined those prior art elements in the normal course of research and development to yield the claimed invention.  Id. at 421"

• "A person of ordinary skill at the time of the invention interprets the prior art using common sense and appropriate perspective.  KSR, 550 U.S. at 421."

The opinion cites the portion of the KSR opinion regarding the relationship between what is "obvious to try" and what is obvious:

When there is a design need or market pressure to solve a problem and there are a finite number of identified, predictable solutions, a person of ordinary skill has good reason to pursue the known options within his or her technical grasp.  If this leads to the anticipated success, it is likely the product not of innovation but of ordinary skill.

The opinion notes that "[a]ccordingly, when design need and market pressure may dictate a commonsensical path using a finite number of identified predictable solutions to one of ordinary skill, deviations from that path are likely products of innovation."  In addition, the opinion notes that the "teaching-suggestion-motivation (TSM)" test provides "one way to identify 'sources of evidence that an ordinary skilled artisan might have found and combined at the time of the invention,'" citing Ortho-McNeil Pharm., Inc. v. Mylan Labs., Inc., 520 F.3d 1358, 1364-65 (Fed. Cir. 2008).

Applying these principles, the Court held that "the patent claims a new composition or formulation to deliver an FDA-approved active ingredient.  Thus, the claimed invention is not obvious if a person of ordinary skill would not select and combine the prior art references to reach the claimed composition or formulation," citing Eli Lilly v. Zenith Goldline Pharm., 471 F.3d 1369, 1380 (Fed. Cir. 2006).  Under the circumstances of this case (a reformulation of a known formulation of a known compound), the opinion states that "[a] prima facie case of obviousness in the chemical arts is often based on a known compound, called a 'lead compound,' which serves as a starting point for a person of ordinary skill developing the claimed invention," citing Eisai Co. Ltd. v. Dr. Reddy's Labs., 533 F.3d 1353, 1357 (Fed. Cir. 2008).  Such a lead compound then provides a basis to show "structural similarities" between the prior art and the claimed lead compound.  But for formulation cases "where the patented formulation was made to mimic a previously FDA-approved formulation," the Court says the term "reference composition" is a better way to frame the issue, since the "functional and pharmaceutical properties" of the composition may be more informative ("appropriate") than the chemical structure of the lead compound.  The Court then looks to the comparison between the Norvartis Miacalcin® product, which the Court finds is the substitution of 20 mM citric acid for BZK.

The Court recognized that the art provided "design need and market demand" for a nasal formulation of calcitonin, identifying the "design need" as the bioequivalent formulation, and the market demand for a composition that treats the same symptoms.  The Court's opinion then focused on the citric acid component -- in part from statements made at oral argument -- and says that "the inclusion of 'about 20 mM citric acid' in the composition provides the strongest case for nonobviousness."

Prior art references considered by the District Court and relevant to the panel opinion include U.S. Patent No. 5,912,014 (which named the inventor of the patent-in-suit as an inventor), which disclosed an oral calcitonin formulation comprising enteric coating and further disclosed experiments on the effects of citric acid on bioavailability and absorption in vivo (albeit showing but minor effects).  With regard to this reference, the opinion states that "this court agrees that no reasonable juror could conclude that the '014 patent would give a person of ordinary skill sufficient reason or motivation to use about 20 mM citric acid in a liquid nasal salmon calcitonin composition."  The Court's reasoning distinguished the formulation claimed in the '812 reissue patent from the oral formulations disclosed in the '014 patent on the difference between an oral dosage form and a liquid formulation, and the "significant" differences in route of administration and formulation.  These facts "would not cause a person of ordinary skill to replace BZK in Miacalcin® with 20 mM of citric acid in the normal course of research and development" in the Court's opinion.  In addition, other prior art referenced -- including U.S. Patent Nos. 5,124,315 and 4,476,116 -- showed only a "vague role" played by citric acid in formulations disclosed in those patents.  The Court cited the '116 patent as teaching from using citric acid as disclosed in the '812 reissue patent, because the '116 patent:

[L]ists over fifty examples, including citric acid, of pharmaceutically acceptable chelating agents to serve as absorption agents . . . .  Both parties agree that the '315 patent reports that the compounds listed in the '116 patent yielded "discouraging" test results, and that "only ammonium tartrate is a satisfactory stabilizing agent for liquid nasal compositions containing polypeptides as active ingredient [sic]."  . . .  One of ordinary skill in the art reading the '315 and '116 patents would have considered about 20 mM citric acid undesirable in a liquid nasal formulation containing salmon calcitonin."

Yet another prior art reference, the Day reference, lists benzyl alcohol and phenylethyl alcohol, two other components of the claimed formulation, as "Excipients used in aqueous nasal products."  The Court characterized this reference as follows:

BZK is one of the nine listed preservatives in Day, along with benzethonium chloride, chlorobutanol, methylparaben, phenylmercuric acetate, propylparaben, and thimerosal.  Citric acid is not included in the list of preservatives, but appears instead as a pH adjuster or buffer.  The Day reference also lists polysorbate 20 and 80 as one of three surfactants used as excipients in aqueous nasal products.  With reference to this prior art, there is no evidence to support the conclusion that a person of ordinary skill would expect a combination of citric acid, benzyl alcohol, phenylethyl alcohol, and polysorbate 80 to contain a buffer, pH adjuster, preservative, and surfactant, but no absorption enhancer or excipient to promote bioavailability.

Finally, the opinion states that the affirmatively recited limitation "about 20.0 mM citric acid" "alone supports" the District Court's nonobvious determination:

When used as an absorption enhancer in the '116 patent, citric acid was one of over fifty options.  See KSR, 550 U.S. at 421.  Further, when the prior art used citric acid at about 20 mM, as in the '315 patent, it was used only as a buffer.  There is no genuine dispute of material fact that a person of ordinary skill attempting to make a liquid composition to deliver salmon calcitonin into a human body through nasal administration, would not have considered using about 20 mM citric acid with the narrowly claimed amounts of benzyl alcohol, phenylethyl alcohol, and polysorbate 80, because the formulation would not be expected to perform properly to meet the specificity of a pharmaceutical use.  Thus, even accepting that there was a design need and market pressure to develop a pharmaceutical formulation that is bioequivalent to Miacalcin®, there is no evidence in the record that claim 19 would be an obvious solution to those motivations.

This outcome is in contrast with the decision in Pfizer v. Apotex, issued in the shadow of the (then) impending KSR decision.  That case involved a novel formulation of a known compound, amlodipine, as a besylate salt, sold under the trademark Norvasc® for treating hypertension and chronic stable and vasospastic angina.  The patent in the Pfizer case, U.S. Patent No. 4,879,303, claimed the besylate salt, while prior art (U.S. Patent No. 4,572,909) disclosed the maleate salt of the active ingredient, 2-[(2-aminoethoxy)methyl]-
4-(2-chlorophenyl)-3-ethoxycarbonyl-5-methoxycarbonyl-6-methyl-1,4-dihydropyridine.  In Pfizer, the Federal Circuit reversed a District Court finding of nonobviousness (determined under the TSM test), relying in part on a reference (Berge) that disclosed 53 FDA-approved anions for making pharmaceutically-acceptable salts, including the anion -- benzene sulphonate -- used to make amlodipine besylate.  There, the choice of salt was considered merely to reflect a design choice and the beneficial and purportedly unexpected properties were discounted by the panel.  In today's opinion, the Court appears to have returned to its prior tendency to understand that the pharmaceutical formulation arts are sufficiently unpredictable that there is a real threat of hindsight reconstruction in the face of successful novel formulations of known compounds.  As Judge Rader stated in his dissent to the Court's refusal to rehear the Pfizer case en banc, "[w]ith unpredictable pharmaceutical inventions, this court more wisely employs a reasonable expectation of success analysis," and since salt selection is unpredictable, there would be no reasonable expectation of success (as three District Court judges had previously found).

Thus, the Federal Circuit appears to have absorbed the thrust of the Supreme Court's KSR opinion, to give more credence to the skilled worker's capacity to recognize beneficial combinations of the prior art, and to consider whether the totality of the teachings of the art support the conclusion that even if it would be obvious to try, a novel formulation is "not obvious if a person of ordinary skill would not select and combine the prior art references to reach the claimed composition or formulation."

Unigene Labs., Inc. v. Apotex, Inc. (Fed. Cir. 2011)
Panel: Cheif Judge Rader and Circuit Judges Moore and O'Malley
Opinion by Chief Judge Rader

Comments

[EG]

Kevin,

A "breath of fresh" air from the Federal Circuit on how to manage the nebulous, almost unusable guidelines in KSR International on how to determine obviousness. If only the Federal Circuit could be consistent on this subject.

[Brad Crawford]

The Pfizer v. Apotex decision was wrong, plain and simple. I am glad to see that the Fed. Cir. appears to be takine a more reasoned approach to chemical obviousness.

[General Admission]

To me, these 103 cases demonstrate how critical presentation of the facts/evidence is in chem/bio cases. Recall that Rader had a hand in the In re Kubin debacle, and (I believe) even said that biochemistry is more like computer science than chemistry. *shudder*

[Generic Manufacturers are Bad for America]

Awesome victory for Big Pharma!!!!!

[Don't Stop the Shilling]

"In Pfizer, the Federal Circuit reversed a District Court finding of nonobviousness (determined under the TSM test), relying in part on a reference (Berge) that disclosed 53 FDA-approved anions for making pharmaceutically-acceptable salts, including the anion -- benzene sulphonate -- used to make amlodipine besylate. There, the choice of salt was considered merely to reflect a design choice and the beneficial and purportedly unexpected properties were discounted by the panel. In today's opinion, the Court appears to have returned to its prior tendency to understand that the pharmaceutical formulation arts are sufficiently unpredictable that there is a real threat of hindsight reconstruction in the face of successful novel formulations of known compounds."

Or the facts were simply better for the patentee in this case.

Pfizer seemed (and still seems) like a very reasonable decision.

[Don't Stop the Shilling]

"these 103 cases demonstrate how critical presentation of the facts/evidence is in chem/bio cases."

This is true of all litigated cases where a great deal is at stake and both parties are invested in the result.

"Recall that Rader had a hand in the In re Kubin debacle, and (I believe) even said that biochemistry is more like computer science than chemistry."

Oh yes, the "debacle" that destroyed the biotechnology industry overnight. Who can ever forget it?

[Dr. Mayur Sankalia]

It is really very interesting to know about the nonobviousness for Fortical. As a formulation scientist, I would like to state that we can only make a different formulation compositions but the real thing is the language (used to construct claim in patent), its interpretation and supportive arguments/cases which make the composition obvious or non-obvious.

[Sean]

Kevin and all, I am in-house IP counsel for Upsher-Smith (the co-plaintiff in this case), and I don't want to comment on this case for obvious reasons, but I will comment on Pfizer.

Pfizer was wrongly decided, by a court trying to save face in view of KSR. (One bit of supporting evidence is that the CAFC invalidated the patent THREE DAYS before its natural expiration. Why waste ink on an opinion in that situation?)

Steve Berge, the author of the notorious Berge reference, sits across the hall from me. I brought the Pfizer decision to his attention when it first came out. He was shocked that they would find the besylate salt obvious, because salts are inherently unpredictable. He was not pleased that his reference was actually used to support the contrary proposition. Frankly, I am surprised that no one from Pfizer contacted him for testimony. (But then again, they had three district court opinions in their favor.)

Pfizer was wrongly decided and should have been overturned en banc. It stands for the proposition that salts are not patentably distinct, regardless of secondary considerations. In short, the rationale to make any pharmaceutical salt is (supposedly) provided by Berge. That cannot be the right result. So any decisions that the CAFC makes now to distance itself from Pfizer are to the good.

[Paul Cole]

It should never be forgotten that cases are fact specific.

As I remember in Pfizer there was a short list of about 7 anions with a realistic chance of success, amongst which besylate was prominent. That may have been true or false as a matter of technical reality, but it is what had been established on the trial record. Apparently (and not going too far into the specifics) this case was different.

There has been a run of CAFC decisions recently where rejections based on a combination of hindsight-selected references has been refuted. Some of these reference combinations could only realistically have been created using the infinite improbability drive of the Heart of Gold spaceship from the Hitchhiker's Guide to the Galaxy, see particularly In re Klein. If commonsense is to be applied in validity determinations then it should be exercised in the selection of the references on which rejections are based. Over-exuberant examiners in the USPTO and elsewhere should take note.

[Sean]

Paul, what does "chance of success" mean? Success with respect to what? I'm not trying to be argumentative. What I'm getting at is that the maleate salt was suitable (i.e., "successful") under some criteria, so why would one go beyond that? Recall, in addition to Pfizer's development of amlodipine maleate, DRL tried to develop (and maybe did develop) a maleate version, although it was never marketed.

The besylate solved problems OTHER THAN those that were outlined in the specification or the earlier patent, and problems that were likely known only to Pfizer at the time of filing. In short, the CAFC used Pfizer's own knowledge to establish its motivation to look for a different salt form. In addition, they used Pfizer's own knowledge and testimony to establish reasonable expectation of success. There is (to my knowledge) nothing in the prior art that indicated the problems with the maleate salt, nor anything that would have specifically indicated "success" in overcoming those specific problems (known only to Pfizer) by switching to besylate.

There are plenty of indicia of nonobviousness. In the Pfizer case, the appellate judges substituted their view for the PTO and for three district court judges who found nonobviousness over the same prior art. The CAFC panel went so far as to say that it was nearly an unrebuttable prima facie case. So to save face for themselves in view of the impending KSR decision, they had to denigrate the prior decisions of three judges and at least one examiner.

[Paul Cole]

@Sean

Courts do not decide cases on objective technical reality, but as I have said, on the record before them. A court is essentially a machine for evaluating testimony in the light of the arguments of counsel. Here are the key quotes from Pfizer

"Dr. Wells readily compiled a list of seven alternative anions—including the besylate—each of which he expected would form an amlodipine acid addition salt:

Q. And one of the reasons why you chose these various salts [sic], or suggested these various salts [sic], is because you expected that they would be able to make a salt of them, correct?

A. There was an expectation, but that wasn’t guaranteed.

Q. And when you chose these salts . . . you believed that if you could, in fact, make an amlodipine salt out of them, these might be a cure for the problems you were having with maleate, correct?

A. Indeed."

With this testimony, no doubt elicited by skilled and well-prepared counsel, there was clear evidence on the record that besylate was obvious and that its advantages was what we Europeans would describe as a mere bonus effect.

I fully accept that there was politics at the relevant time, and that the “choice from a few logical alternatives” language in KSR was inspired by Pfizer. But that case was decided according to the trial testimony. If you are in any doubt about that have a look at the little gem of trial testimony at footnote 13 of Graham v John Deere:

"Q. . . . Do you regard the small degree of flex in the forward end of the shank that lies between the pivot point and the point of spring attachment to be of any significance or any importance to the functioning of a device such as 798?

A. Unless you are approaching the elastic limit, I think this flexing will reduce the maximum stress at the point of pivot there, where the maximum stress does occur. I think it will reduce that. I don't know how much.

Q. Do you think it is a substantial factor, a factor of importance in the functioning of the structure?"

"A. Not a great factor, no."

With that testimony on the record, counsel for Graham must have just wanted a dark place within which to hide and nurse his grief: the case had been lost in an instant. However well you prepare and however meritorious the case, it can be lost in a few moments of testimony, and that is a reality that we all face.

[Paul Cole]

On reflection, Pfizer was lost on a single word: "indeed".

If Dr Wells had said that although you might get a salt there was no way of knowing in advance what its properties would be, whether it would be better or worse than maleate, or whether a salt with better properties than maleate even existed, then the decision should have been otherwise. Scientific caution would have suggested this view. But he went along with counsel, probably without realizing the significance of what he had said, and the decision followed.

You can discuss a case at length with a witness. You can ask him the questions the other side might ask, and ascertain what the answers are likely to be. But when the witness goes into the box and is asked questions by opposing counsel it is like a child's paper boat launched on a pond: the wind will blow it where it will.

It is exactly the same before the UK courts because our two legal systems are in essential respects the same. And it is no different in the EPO: if you have a witness testifying or if a technical expert is asked a question by a member of the examining or opposition division or appeal board, the case can be lost (or equally possibly won) on a word.

[Sean]

@Paul, you seem to be slipping into the same trap as the Pfizer CAFC panel, which is to treat the inventor's testimony as determinative of what was obvious to the person skilled in the art. There is no way Wells' testimony regarding how he arrived at the invention should be used against Pfizer (see 103(a)).

I will note that Rader's opinion in Unigene is careful not to confuse Dr. Stern with the POSA -- see page 18.

Had Dr. Berge (who would be an expert who could testify as to the knowledge of a POSA) been asked, he would have stated exactly what you wrote -- there was no way of knowing in advance what the salt's properties would be, whether it would be better or worse than maleate, or whether a salt with better properties than maleate even existed.

But in the end, this case was won 3x over by Pfizer, and only lost because the CAFC decided they needed to save face. The testimony/evidence had little to do with it. As many judges do, they cherry-picked evidence to support their decision, rather than taking the evidence as a whole and providing some deference to the numerous lower decisionmakers.

[Paul Cole]

@ Sean

There is much case law in Europe that a court or the EPO can disregard inventor testimony where the inventor is not representative of a person skilled in the art. But I know of no authority that a court must disbelieve an inventor's testimony on inventive step where that testimony is credible and considered to be representative of the prior art. The mere fact that Dr Stern was not considered to be typical in Unigene is no reason for saying that the court was wrong for believing Dr Wells in Pfizer. It is all down to the individuals and the trial transcript.

Furthermore, this is surely an area where there is some flexibility available to the judges. If there is no rigid rule, there can be no "trap".