Extension survives, adjustment does not

How to read the ledger below. Each numbered point carries the kind of evidence behind it. ESTABLISHED is stated in the authority's own text, with a locator. IMPLEMENTED is the rule as the authority actually applied it. OBSERVED is carried by a source outside this one. ASSUMED is the drafter's inference, flagged as such and not the authority speaking. Quotations are checked against a cached copy of the primary source by script.

In one paragraph

The Federal Circuit affirmed the District of Delaware in Novartis v. Ezra Ventures on 7 December 2018 and held that obviousness-type double patenting does not invalidate a validly obtained patent term extension under Section 156, so long as the patent was already valid at its expiration date before the extension was added. The court also held that extending one patent does not breach the one-patent limit merely because it effectively extends a related patent, because the limit governs the legal status the Director confers and the patentee chooses which patent receives it.

Reader summary

Novartis owned two patents covering fingolimod: the '229 patent, a pre-URAA patent claiming the compound, and the '565 patent, a post-URAA patent claiming a method of administering it. Novartis chose the '229 patent for a Hatch-Waxman term extension, which pushed its expiration past the '565 patent's. Ezra, a generic filer, argued two things: that extending the '229 patent also "effectively" extended the '565 patent and so broke the statutory rule that only one patent may be extended; and that the '229 patent was invalid for obviousness-type double patenting over the '565 patent because it now expired later.

Both arguments failed. On the statute, the court refused to read "effectively" into a provision that does not contain it. The one-patent limit is about the legal status the Director confers, not about the practical consequence for related patents, and the patentee picks which patent gets it.

On double patenting, the court extended Merck v. Hi-Tech and held that the doctrine does not invalidate a validly obtained extension. But the holding carries a condition that is easy to drop when the case is summarized, and the condition is the operative part: the extension is safe only if the patent was already valid at its pre-extension expiration date. A patent that should have been terminally disclaimed back then can still be invalidated, and the extension will not rescue it.

The court also explained why the usual concern was absent. Here the earlier-filed, earlier-issued patent was the one expiring later, and only because of a statutory extension. That sentence was later quoted by Allergan v. MSN. Five years later In re Cellect drew the opposite conclusion for patent term adjustment, and the two cases together are why the distinction between the two mechanisms matters.

Why this matters when you are drafting

The headline is only half the rule. "Extension survives double patenting" is how this case is usually summarized, and that summary is wrong in the way that matters. Claim 4 is the rule: the extension survives because the patent was already valid at its pre-extension expiration date. If a terminal disclaimer should have been filed and was not, the extension changes nothing.

This is the other half of the Cellect pair, and the pairing is the point. Extension under Section 156 is added after the double-patenting comparison; adjustment under Section 154 is added before it. Same family, same arithmetic, opposite result, and the difference is which statute conferred the extra time. When you audit a family's term exposure, the first question is not how much extra term exists but which mechanism produced it.

Do the counterfactual, not the observation. Claim 5 is the operational instruction. Comparing today's expiration dates tells you nothing here, because the extension is exactly what moved them. Strip the extension, then compare.

The choice of which patent to extend is yours, and it is strategic. Claims 8 and 9 mean the patentee selects, and a de facto benefit to relatives is a permissible consequence rather than a statutory violation. That is a genuine planning lever at the point a family faces regulatory review.

Mind the regime boundary. Claim 10 flags a pre-URAA and post-URAA pair, which this corpus treats separately in its effective-filing-date tier. If a family straddles the 1995 change, the term rules are not uniform across its members, and the court itself pointed to a companion case for that problem.

What the authority establishes

[ESTABLISHED, the disposition] Affirmed, and specifically: "the district court was correct in finding that the '565 patent is not a double patenting reference to the '229 patent". (P. 14.)

[ESTABLISHED, the holding as the court states it] "a PTE pursuant to § 156 is valid so long as the extended patent is otherwise valid without the extension". (P. 14.)

[ESTABLISHED, the same holding stated earlier] "obviousness-type double patenting does not invalidate a validly obtained PTE in such a scenario", reached "as a logical extension" of Merck v. Hi-Tech. (P. 10.)

[ESTABLISHED, and this is the condition summaries drop] The rule has two branches. "if a patent, under its original expiration date without a PTE, should have been (but was not) terminally disclaimed because of obviousness-type double patenting, then this court's obviousness-type double patenting case law would apply, and the patent could be invalidated". (P. 12.) Only then the safe branch: "if a patent, under its pre-PTE expiration date, is valid under all other provisions of law, then it is entitled to the full term of its PTE". (P. 12.)

[IMPLEMENTED] The operative test is therefore counterfactual, not observational. You do not ask whether the patent as extended expires later than a relative; you ask whether the patent at its pre-extension expiration date was already exposed. The court applies exactly that: but for the extension, the '229 patent would have expired before the '565 patent. (Pp. 12 to 13.)

[ESTABLISHED, why the usual concern was absent] "This case does not raise the traditional concern with obviousness-type double patenting" (P. 12.), because "it is the earlier-filed, earlier-issued '229 patent, not the later-filed, later-issued '565 patent, that has the later expiration date". (P. 12.) That sentence is the one Allergan v. MSN later quotes.

[ESTABLISHED, the policy line] "agreeing with Ezra would mean that a judge-made doctrine would cut off a statutorily-authorized time extension. We decline to do so." (P. 13.)

[ESTABLISHED, the one-patent limit is about legal status, not effect] "there is no reason to read 'effectively' as a modifier to 'extend' in the language of § 156(c)(4)" (P. 8.), and the provision "was intended to limit a legally conferred PTE (not an 'effective' or 'de facto' PTE) to one patent selected by the patent owner". (P. 9.)

[ESTABLISHED, the patentee chooses] "nothing in the statute restricts the patent owner's choice for patent term extension" (P. 8.), and quoting Merck, "Congress chose not to limit the availability of a patent term extension to a specific patent and instead chose 'a flexible approach which gave the patentee the choice'". (P. 9.)

[ESTABLISHED, the two patents sat in different term regimes] "The '229 patent is thus a pre-URAA patent whereas the '565 patent is a post-URAA patent, governed by different statutory patent term regimes." (P. 4.) The court notes in footnote 3 that a mixed pre-URAA and post-URAA pair can raise its own issues, pointing to Breckenridge, decided the same day. (P. 13 n.3.)

[ESTABLISHED, Gilead is distinguished on gamesmanship, not on principle] Gilead addressed inventors who could "routinely orchestrate" longer exclusivity by structuring priority claims and issue order; here "Ezra does not identify any similar tactics on the part of Novartis". (P. 13.) Footnote 2 adds that "The effect of statutory term extensions was expressly not considered in Gilead". (P. 13 n.2.)

[ESTABLISHED, what the court did not decide] Because the '565 patent was not a proper reference, "we need not address Ezra's arguments as to whether the '229 patent is patentably indistinct from the '565 patent". (P. 14 n.4.) The opinion therefore contains no holding on whether these claims were in fact patentably indistinct.

[ASSUMED, drafter's inference, flagged as such] Claims 2 and 4 read together mean the extension is not a cure but a preservation: it protects a position that was already sound. A practitioner treating an extension as a way to outrun a double-patenting problem has the causation backwards. The opinion does not put it this way; it follows from the conditional structure of the holding.

Quotations, verified against the source

All from the Federal Circuit's opinion. Page locators computed from the running header, since this 2018 opinion predates the docket stamp.

"The '229 patent is thus a pre-URAA patent whereas the '565 patent is a post-URAA patent, governed by different statutory patent term regimes" (P. 4.)
"nothing in the statute restricts the patent owner's choice for patent term extension" (P. 8.)
"there is no reason to read 'effectively' as a modifier to 'extend' in the language of § 156(c)(4)" (P. 8.)
"was intended to limit a legally conferred PTE (not an 'effective' or 'de facto' PTE) to one patent selected by the patent owner" (P. 9.)
"obviousness-type double patenting does not invalidate a validly obtained PTE in such a scenario" (P. 10.)
"This case does not raise the traditional concern with obviousness-type double patenting" (P. 12.)
"it is the earlier-filed, earlier-issued '229 patent, not the later-filed, later-issued '565 patent, that has the later expiration date" (P. 12.)
"if a patent, under its pre-PTE expiration date, is valid under all other provisions of law, then it is entitled to the full term of its PTE" (P. 12.)
"agreeing with Ezra would mean that a judge-made doctrine would cut off a statutorily-authorized time extension. We decline to do so" (P. 13.)
"The effect of statutory term extensions was expressly not considered in Gilead" (P. 13.)
"a PTE pursuant to § 156 is valid so long as the extended patent is otherwise valid without the extension" (P. 14.)
"the district court was correct in finding that the '565 patent is not a double patenting reference to the '229 patent" (P. 14.)

Tensions and open questions

"Otherwise valid without the extension" is stated, not operationalized. Claim 2 gives the standard and claim 4 gives its two branches, but the opinion does not say who bears the burden of showing the pre-extension position was sound, or at what point in a proceeding it is tested.

The de jure and de facto distinction is doing heavy work. Claim 8 concedes the practical effect Ezra complained of, that the related patent's subject matter cannot be practiced during the extended term, and calls it "a permissible consequence." Whether that reasoning survives a case where the practical effect is larger is untested here.

The relationship to Cellect is not stated in this opinion, and could not be. Cellect came five years later and reached the opposite result for adjustment. Reading this case alone will leave a practitioner with a rule that is correct for extension and dangerous if carried across to adjustment.

No holding on patentable indistinctness. Claim 12: the court expressly declined it. This record cannot support any proposition about whether the two patents' claims were in fact indistinct.

What this record does and does not cover

The opinion was read end to end, all 14 pages, including both Discussion parts, the conclusion and all four footnotes. Footnotes 2, 3 and 4 are in the ledger because each limits the holding in a way the body does not.

What was NOT read closely, and is therefore outside what this record can support: Merck v. Hi-Tech, Gilead and Breckenridge as separate authorities. This record states what this court said about them, not what they hold. The district court opinion and the patents themselves are known here only through this court's characterization. The page 4 timeline figure was not recoverable from the text extraction, so any claim about the precise interval between the two patents' expiration dates beyond the dates quoted is outside this record.

Currency

This record was verified against its primary source on 2026-08-15. That is a fidelity check: the quotations, locators and attribution are accurate to that source. Whether the authority has since been narrowed, distinguished, abrogated or reversed has not been checked here. Confirm it still stands before relying on it.