What Were They Thinking? A Second Look at Lilly v. Sun
As we all learned years ago, when trying to make sense of Lilly v. Barr, if a Fed. Cir. decision wrestles with obviousness-type double patenting, it will be a labor of Hercules to reason it out. But the majority of the Fed. Cir. really blew it in Eli Lilly v. Sun Pharma. Ind., 611 F.3d 1381 (Fed. Cir. 2010), both in the original decision and in denying en banc review. Newman, Rader, Lurie and Linn made up the minority in that vote, and when those four agree on something, attention must be paid.
But I don’t feel that much attention was paid. Maybe the decision was too close to the holidays. Maybe folks just thought, “Just another case of Lilly trying to evergreen the patent protection on a blockbuster drug [gemcitabine].” I got re-interested in the decision when I read the amicus brief filed by the Washington Legal Foundation in support of Lilly’s petition to the Supreme Court to grant cert. Although I have dipped my toe in the sometimes murky waters of this right-wing think tank in the past- I summarized the Supreme Court’s decision in Pioneer Hi-Bred v. JEM – they are usually more interested in supporting suits to reverse “Obamacare” or to i.d. illegal immigrants. Whatever your political leanings, its amicus brief was well-written (even if it never used the term “obviousness-type double patenting” and framed the question too broadly). In any case, it got me to re-consider just how bad the panel decision was.
It is a little hard to follow the timeline of the filings of the two patents in question without a timeline, but I will try. To begin with, both patents issued out of pre-GATT filings and so got 17 years from their issue dates. The “oldest patent application” was filed in 1983 claiming gemcitabine (“gem”) and its use as an anti-viral (Of course this was at the beginning of the AIDS epidemic). Lilly subsequently discovered that gem was an effective anti-cancer drug (now Gemzar) but, instead of just filing a CIP of its pending application disclosing, and perhaps claiming, the new use, they ALSO filed an original application on the same day (Dec. 4, 1984) claimed the new use. The “oldest patent application” claiming gem issued as the ’614 patent on Feb. 28, 1989 and Lilly got PTE, so that it did not expire until May 15, 2010. Lilly did not file a divisional on the use of gem to treat cancer, although this would have been a pre-GATT filing and given Lilly a lot of extra patent term.
Instead, Lilly simply prosecuted the second application until it issued on November 11, 1995, claiming the use of gem to treat cancer. This pre-GATT filing will not expire until November 7, 2012. Lilly thought it was all set until then. But Sun got the Fed. Cir. to find the second ’826 patent invalid for obviousness-type double patenting on the basis that the claimed use was disclosed – but never claimed – in the soon to expire ‘614 patent. Somehow the panel found that Lilly’s approach led it into the deadly swamp of obviousness-type double patenting, and would be an improper extension of patent term to let the ‘826 patent expire later than the ‘614 patent.
I am not going to argue the panel’s position. I can’t. But let’s take a quick look at Lilly alternatives. The best one would have been to have never filed the second “cancer” application. The CIP with claims to gem and to two distinct uses would surely have drawn a restriction requirement. If Lilly simply obtained the ‘614 patent to gem and then filed a divisional to treating cancer in 1989, the second patent would probably have had the same term as the invalidated ‘826 patent (if it issued on about the same date, November 7, 1995).(Note that it does appear that Lilly might have slowed down the prosecution of the ‘826 patent to get maximum term, but Lilly could have done the same thing with the hypothetical divisional.) The divisional would not have be subjected to an obviousness type double patenting rejection. End of story.
Lilly had yet another option. Lilly could have not filed a CIP at all (Lilly got nothing useful out of it), but simply obtained the ‘614 patent to gem based on its anti-viral utility. Lilly could have still filed the second application disclosing the cancer treatment in 1984 and prosecuted it to obtain the ‘826 patent. Now there would be no disclosure of the cancer treatment in the ‘614 gem patent, and no basis on which to find the ‘826 patent “guilty” of obviousness-type double patenting.
I can’t speculate on why Lilly chose the course it did, but if 4 of the most scientifically-learned Fed. Cir. judges would not have seen an obviousness-type double patenting defense to infringement coming, Lilly can hardly be blamed. The Washington Legal Foundation framed the question: “Whether the [Fed. Cir.] erred by holding that the mere description of an invention in a patent renders a subsequently-claimed invention invalid.” That was not the holding below, but WLF was on point in starting out with a quote from Miller v. Eagle Mfg. Co., 151 U.S. 186 (1984): “[A] later patent may be granted where the invention is clearly distinct from, and independent of, one previously patented.”