Sun Pharmaceutical Industries, Ltd. v. Eli Lilly & Co. (Fed. Cir. 2010)

By Andrew Williams --

On Wednesday, in Sun Pharm. Indus., Ltd. v. Eli Lilly & Co., the Federal Circuit affirmed a finding that claims of Lilly's U.S. Patent No. 5,464,826 ("the '826 patent) were invalid for obviousness-type double patenting over U.S. Patent No. 4,808,614 ("the '614 patent").  The '614 patent claims gemcitabine, and methods of using gemcitabine to treat viral infections, while the '826 patent claims methods of using gemcitabine for treating cancer.  Lilly markets the drug GEMZAR, which contains gemcitabine as the active ingredient, for the treatment of various forms of cancer.  Correspondingly, both the '614 patent and the '826 patent were listed in the Orange Book with respect to GEMZAR.  Sun filed an ANDA in 2006 to market a generic version of GEMZAR, and filed a declaratory judgment action in 2007, asserting, among other things, that the '826 patent was invalid.  Ultimately, the U.S. District Court for the Eastern District of Michigan granted Sun's motion for partial summary judgment because the asserted claims of the '826 were invalid based on obviousness-type double patenting.

Lilly's '614 patent was a continuation-in-part ("CIP") application of U.S. Application No. 06/473,883 ("the '833 application").  The original '833 application only described gemcitabine's antiviral activity.  However, when the CIP application was filed, the specification was amended to include the observation that "certain compounds of the present invention have also demonstrated excellent oncolytic activity in standard cancer screens" ('614 patent, col. 17, ln. 54-56).  Nevertheless, the '614 patent did not claim any methods of treating cancer.  In fact, Sun and Lilly did not dispute that the '614 patent was entitled to the benefit of the '833 application filing date.  On the same date that Lilly filed the application that resulted in the '614 patent, it also filed an application that gave rise to the '826 patent.  Because both of these applications were filed prior to 1995 (December 4, 1984 to be exact), the patent term was calculated based on the issue date.  Lilly obtained patent term extension for the '614 patent, and as a result, the patent expired on May 15, 2010.  The '826 patent was set to expire two-and-a-half years later, on November 7, 2012.  Lilly did not file a terminal disclaimer.

The Federal Circuit agreed with the District Court in finding that this case was analogous to the prior decisions in Geneva Pharm., Inc. v. GlaxoSmithKline PLC, 349 F.3d 1373 (Fed. Cir. 2003), and Pfizer, Inc. v. Teva Pharm. USA, Inc., 518 F.3d 1353 (Fed. Cir. 2008).  In Geneva Pharm., an earlier patent claimed the compound potassium clavulanate, and the specification disclosed its utility in inhibiting B-lactamase, while a later patent claimed the method of using the compound to inhibit B-lactamase.  Because the second patent did no more than claim the previously disclosed utility of a previously claimed compound, the two patents were not patentably distinct.  In Pfizer, an earlier patent claimed several COX-2 inhibitors, and disclosed that they could be used to treat both inflammation and inflammation-associated disorders.  A later patent claimed these previously disclosed methods of use.  Again, the Court found that these two patents were not patentably distinct.

Lilly asserted two reasons why the present case was different from both Geneva Pharm. and Pfizer.  First, Lilly pointed out that in the earlier cases, there was only one disclosed utility -- the methods of use that were claimed in the later patents.  In contrast, Lilly's earlier patent disclosed more than one utility -- treating viral infections and cancer.  Because the treatment of cancer was not the "essential utility," Lilly argued, the two patents were patentably distinct.  The Court rejected this argument because the Pfizer case, in fact, did involve more than one essential utility -- the treatment of inflammation, and the treatment of inflammation-associated disorders.  Of course, the Court does not address whether these were results of the same utility, namely, whether the treatment of the inflammation-associated disorders would flow naturally from the treatment of the inflammation.  But, on the other hand, it is not clear from the discussion whether the anti-cancer properties of gemcitabine would not naturally result from the treatment of the viral infections.  Nevertheless, the Court determined that the holdings of Geneva Pharm. and Pfizer extend to any and all uses disclosed in the earlier specification.

The Federal Circuit also rejected Lilly's second argument that the trial court erred in consulting the specification of the issued patent instead of the specification of the priority document, the '833 application.  In support of this argument, Lilly relied on claim construction concepts as articulated in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005).  Specifically, Lilly argued that a claim term should be given its ordinary and customary meaning as understood by one of ordinary skill in the art at the time of invention.  Gemcitabine's anti-cancer properties were unknown at the time the '614 patent was invented.  However, the Federal Circuit pointed out that its claim construction precedent establishes that claim terms must be construed in light of the entire issued patent, and therefore found no support for Lilly's argument.

Ultimately, the Federal Circuit appears to justify its decision on the policy considerations as articulated in both Pfizer and Geneva Pharm.:

It would shock one's sense of justice if an inventor could receive a patent upon a composition of matter, setting out at length in the specification the useful purposes of such composition, . . . and then prevent the public from making any beneficial use of such product by securing patent upon each of the uses to which it may be adapted.

Pfizer, 518 F.3d at 1368 n.8; Geneva Pharm., 349 F.3d at 1386.

Sun Pharmaceutical Industries, Ltd. v. Eli Lilly & Co. (Fed. Cir. 2010)
Panel: Circuit Judges Bryson, Gajarsa, and Prost
Opinion by Circuit Judge Prost

Comments

[EG]

Andrew,

In view of the cited precedent, an unremarkable decision.

[Student]

Let me get this straight.

They filed an early application, '883, that said nothing about cancer. On 12/4/84 the filed a CIP, '614, that mentioned cancer but did not claim the method. On the same day they filed '826 claiming the cancer methods. There was no terminal disclaimer for this second application.

So if they had ommitted the cancer disclosure from the '614 application they would have been in the clear? I don't see a reason why it had to be in the CIP at all.

[Andrew Williams]

Dear Student,

I hadn't actually gone back to check when all of the applications were filed, I was relying on the decision for this information. Your message prompted me to double-check the dates. The Federal Circuit opinion correctly identified filing dates, but did not mention that there were intervening application in between the December 4, 1984 applications and those that ultimately issued. Therefore, to set the record straight, the relevant family history (at least according to the Public Pair website) of the '614 patent is that it was filed on 6/4/1987 as a divisional application of App. No. 06/677,146 (which issued as U.S. Patent 4,692,434), which was filed on 12/4/1984 as a CIP of the '883 application. In addition, the '826 patent was filed on 7/26/1994 as a divisional application of 08/099,268, which was filed on 7/29/93 as a continuation application of 07/746,441, which was filed on 8/16/1991 as a divisional application of 07/163,571 (which issued as U.S. Patent 5,061,793), which was filed on 3/3/1988 as a continuation application of 06/786,419, which was filed on 10/10/1985 as a CIP of 06/677,783, which was filed on 12/4/1984 (and was the first application in that chain). I also didn't check what the various applications disclosed, but the Federal Circuit opinion provides the quote from the '614 patent specification.

In any event, yes, I believe you are correct that if they had left out the disclosure regarding the anti-cancer properties of gemcitabine, the '826 patent would not have had a double-patenting problem. And, yes, they did not claim the anti-cancer properties in the '826 patent, or any patent that issued from that family. However, there is a requirement that an applicant disclose the best mode, and the patent drafter may have had that in mind when they included the anti-cancer disclsoure in the specification. It is often difficult using hind-sight reconstruction to second guess why a disclosure was included, or not included, in an application.

Thanks for the comment.

[Andrew Williams]

EG - I agree that the case appears unremarkable, but I don't think Geneva and Pfizer were as clear-cut as the Court suggests. However, the decision in consistent with the stated policy concerns. Thanks for reading the post anyway.

[Gaby L.]

Would it have been possible for Lilly to statutorily disclaim claims 2, 6 and 7 during litigation or as soon as they received the notice letter?

[Andrew Williams]

Gaby L.

You can file a terminal disclaimer in an issued patent as long as the patent to which you are tying the disclaimer has not already expired. Therefore, Lilly could have filed a disclaimer up until May of this year. However, Lilly would be in the same position as they are now - no current patent protection for the '826 claims. And, as this is an ANDA case, there is no past infringement damages to worry about. So, yes, they could have, but it would not have had much of a practical impact. Thanks for the question.

[athome]

Of course, presumably, Eli Lilly could not have claimed priority back to the original '833 application for the cancer claims, but yet that disclosure was used against them in the obviousness analysis as if the cancer material had been filed with the original '833 disclosure. There is a discussion towards the end of the opinion to the effect that it is the spec in the "issued patent" that counts, and not in the original application which was only "an early draft" of the issued patent. Well that "early draft" was an original spec-as filed! Unremarkable. (Assuming you are not Eli Lilly).

[6]

"It would shock one's sense of justice if an inventor could receive a patent upon a composition of matter, setting out at length in the specification the useful purposes of such composition, . . . and then prevent the public from making any beneficial use of such product by securing patent upon each of the uses to which it may be adapted."

Try telling that to software patent advocates...

[Kevin E. Noonan]

Actually, that quote needs to be qualified by the statement "relating to the uses disclosed in the application." It has long been the case that a new use for an old compound can be patented; while the old compound is not itself patented "again," the new use is.

So if the use was new and not disclosed in the earlier application, there may have been (or should have been) a different result.

In addition, the distinction between inflammation and inflammatory disorders is pretty weak - not that I think Lilly's argument was persuasive, but this is the kind of semantic pin-dancing that can only be done by judges with no understanding of the underlying science.

[Gaby L.]

Andrew,

I was not referring to a terminal disclaimer. Instead, I was asking about a statutory disclaimer. See 35 U.S.C. 253 for a discussion on statutory vs terminal disclaimers.

Under 37 CFR 1.321(a) the owner of a patent may disclaim a complete claim or claims of his or her patent. This may result from a lawsuit or because he or she has reason to believe that the claim or claims are too broad or otherwise invalid. If the patent is involved in an interference, see 37 CFR 41.121(a).

As noted above, a statutory disclaimer is a statement in which a patent owner relinquishes legal rights to one or more claims of a patent.

[Sidney J. Russel]

Am I the only one who's bothered by this line of cases? How is it that a non-use-limited composition of matter claim can be patentably indistinct from a novel method of treatment claim for the purposes of obvious-type double patenting? Straight obviousness possibly (depending on the gap in time between the reference patent application/patent and the follow-on application), but double patenting???