The Patent Act has long prohibited someone from patenting an invention that was already "on sale" over a year before the patent application was filed. A prior sale invalidated the patent if the existence of the sale was public, even if what was sold was secret. But Congress amended the Patent Act in 2011 and revised the requirements for obtaining a patent. The amendment added a new clause: The statute now says you can't patent an invention that was already "on sale, or otherwise available to the public
." The U.S. Court of Appeals for the Federal Circuit held in Helsinn Healthcare S.A. v. Teva Pharmaceuticals USA, Inc.
, 855 F.3d 1356 (Fed. Cir. 2017), that this amendment doesn't change the rule that a sale is invalidating even if the details of what was sold are secret. The Supreme Court will review that decision this coming term.
Prior sale defenses often arise in pharmaceutical cases like Helsinn. Drug development is expensive, so developers often outsource manufacturing or distribution. Contracts between these various entities can create "sales," even if the product doesn't make it to the "general" public. In addition, drug development takes years, which can cause delays in patent filing, as the developer waits for clinical data. Delay creates a risk of sales between various partners over a year before patent filing. And these sales become invalidating if the patent holder reports it publicly.
Helsinn's facts fall within that common pattern. While the patent holder (Helsinn) conducted Phase III clinical trials, it contracted to sell the drug to another entity (MGI), who would distribute it after FDA approval. The contract was binding under the UCC -- it included a price and required the patent holder to meet MGI's requirements -- and it was indisputably for the patented drug. Although the parties didn't make the drug's composition public, they announced other contractual terms in a joint press release and in Helsinn's SEC statement.
The legal issue involves pure statutory interpretation. Helsinn no longer challenges its distribution contract was an invalidating "sale" under the prior statute. It instead argues the America Invents Act, passed in 2011, changed the rule. The act's main purpose was actually to change something unrelated to the dispute here -- it converted America from a "first-to-invent" to a "first-to-file" patent system. But that conversion also reorganized various provisions into 35 U.S.C. Section 102(a)(1), which now prohibits granting a patent where "the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public." Helsinn thinks the amended text requires the details of what was sold to be public.
Helsinn's textual argument relies on reading all the listed items consistently. The "on sale" provision is part of a list where every other item -- i.e., patents, publications, public uses -- is public. What's more, the final clause confirms that each of the previous items must involve publicly available information. It is a catch-all for information "otherwise available to the public," which suggests each prior item is likewise available to the public.
This textual analysis makes sense, although some have expressed doubts. One federal circuit judge thought that the text does not require a public sale, reasoning that "to the public" in "otherwise available to the public" modifies only "otherwise available," and not all the other items in the list. Picking up that thread, the parties debate how to parse the grammar of "series modifiers." This overcomplicates matters. The word "otherwise" is superfluous unless it signals that all the prior items are also publicly available.
But this textual analysis does not dictate a win for Helsinn. The text tells us that the "on sale" provision involves some kind of public availability, but it doesn't answer the key question of what must be public. Is it enough that the occurrence of the sale is public? Must the details of what is sold be public? Or does the text require something even less than what the Federal Circuit suggested -- does it simply that the sale be to a member of the public, i.e., someone other than the inventor, regardless of whether others know about the sale? Each interpretation is consistent with the statutory text.
Helsinn points to the term "claimed invention" at the beginning of the list, but this still doesn't resolve the ambiguity. The statute requires that the "claimed invention was ... on sale," and that some aspect of the sale be public. But, again, that does not compel the conclusion that the details of what were sold must be public. The text is consistent with the Federal Circuit's conclusion that a sale is invalidating if its existence was public and what was sold was the claimed invention.
The parties also debate whether another part of the statute resolves the ambiguity, but its relevance is minimal. Section 102(b)(1) creates an exception for certain "disclosures" that would otherwise be invalidating under Section 102(a)(1), including situations where the inventor made a "public disclosure" less than a year before applying for her patent. One Federal Circuit judge suggested that Section 102(b)(1)'s use of the broader term "disclosure" to describe items falling within Section 102(a)(1) means that Section 102(a)(1) is not limited to "public" disclosures. Indeed, Section 102(b)(1) uses both "disclosure" and "public disclosure" in various places, which the judge thought meant the term "disclosure" isn't by itself limited to public disclosures. But this is a little too clever. Congress was likely just imprecise in using the terms "disclosure" and "public disclosure," and it's hard to believe it intended them to communicate something about the meaning of "on sale" in a different part of the statute. Anyway, the text of Section 102(a)(1) lists only public disclosures for the reasons given above, so Section 102(b)(1)'s generic reference to "disclosures" is necessarily limited to public disclosures because it could include only the public events listed in Section 102(a)(1).
The legislative history is more pertinent to the dispute but also ambiguous. The House and Senate Committee Reports both emphasize that the "otherwise available to the public" language was added to Section 102(a)(1) to "clarify the broad scope of relevant prior art, as well as to emphasize the fact that it must be publicly accessible." You could read this statement to support either side. The first half supports the Federal Circuit's decision by emphasizing the "broad scope" of Section 102(a)(1). The second half might be thought to favor Helsinn, because it confirms that all prior art under Section 102(a)(1) must be "publicly accessible," including items "on sale." But, again, it's unclear what that means -- it doesn't resolve what about the sale must be public. If anything, it implies that the sale must simply be to a member of the public, regardless of whether the details of what was sold (or even that there was a sale), are publicly known. The product would be "accessible" to someone in the "public" if it were sold to anyone other than the inventor, even if the buyer didn't know the details of the product's composition. Diet Coke is accessible to me, even though I don't know how to make it myself.
Several individual members of Congress made floor statements that are even more pertinent, although it's debatable how much weight we should give them. Senator Patrick Leahy said that Section 102(a)(1) "was drafted in part to do away with precedent under current law that private offers for sale ... may be deemed patent-defeating prior art." And Sen. Jon Kyl said that "a secret sale or offer for sale" that "does not operate to disclose the invention to the public" is not invalidating under the new Section 102(a)(1). These senators wanted to change "on sale" law in some way, and Sen. Kyl's remark suggests that the details of what was sold that must be public.
But there is reason to discount this evidence. Prior versions of the bill omitted the "on sale" provision entirely, and it was added back over Sen. Kyl's objection. So his views may not reflect what Congress as a whole thought. The final text used the same "on sale" language from the prior statute, and suggesting that adding the "otherwise available to the public" clause was meant to change the meaning of the items before it is awfully subtle. If Congress really wanted to overrule precedent, why not change the "on sale" language to something that actually set forth a new rule?
The various policies behind the on-sale bar don't answer the question either, because you can argue them either way. The bar prevents inventors from extending their monopolies by delaying their patent filings. But it's debatable whether that happened here. Helsinn didn't start selling to the general public (i.e., patients) until around when it filed for its patent, yet it did sell to someone else (MGI) years earlier, and MGI is a member of the "public" in some sense. Backing up, you might say that we should be more permissive of drug developers who "extend their monopoly," because they often lose effective patent term due to delays in FDA approval. But, if you do that, you also delay the date that generics can enter the market, which may overcompensate the developer and harm the public interest in affordable drugs, depending on the circumstances.
Other considerations point in opposite directions. The on-sale bar is meant to encourage prompt filing to make the patent disclosure available to others who seek to improve upon it. That favors the Federal Circuit's rule. But the on-sale bar also comes with a statutory grace period, which is meant to allow inventors time to test the commercial value of their technology before incurring the expense of a patent application. You could say Helsinn abused that privilege here, as its agreement with MGI already indicated the invention's value but was well outside the one-year grace period. Or you could say that Helsinn didn't really get to test the invention until FDA approval allowed it to sell to the general public, well within the grace period.
Perhaps the only thing this case makes clear is that statutory interpretation can be hard. We often hear that a statute's "plain meaning" resolves the case, that objective canons of construction will lead us to the right result. Not so here. We also hear that legislative history shouldn't be considered at all. But the legislative history here seems more relevant than the text itself, because it at least addresses the dispute in this case. Yet the legislative history here also suffers from the very problems that led to its disfavor -- it provides nuggets that can be argued either way, and it's not clear that it reflects the intent of Congress as a whole.
So there's no objectively "right" answer here, but given all the ambiguity, the Federal Circuit's approach seems like the wise one. The term "on sale" was unchanged by the America Invents Act, so it's sensible to say that it retains the same meaning. If Congress wanted to change the rule, it should've said so expressly, rather than leaving us to guess at its intent and make multiple logical leaps to divine it. It's also notable that the term "on sale" includes offers for sale, indicating that it is not critical that the product actually reach the hands of the general public for a transaction to be invalidating.
Inventors have nothing legitimate to fear from the Federal Circuit's rule. There are protections in existing law that make the Federal Circuit's interpretation of the on-sale bar inapplicable to transactions where the inventor is still experimenting with its product or where the inventor doesn't yet know whether its product will work. And, of course, the inventor controls its patent filing. Inventors can avoid any issues with the on sale bar by simply filing early and erring on the side of caution.