Burroughs Wellcome v. Barr: Patents

The Federal Circuit's conception standard: inventorship fixes when the idea is definite and permanent, so NIH scientists confirming AZT were not co-inventors.

Burroughs Wellcome Co. v. Barr Laboratories, Inc., 40 F.3d 1223 (Fed. Cir. 1994), decided November 22, 1994, is the Federal Circuit’s leading statement on the meaning of conception, the doctrinal anchor of inventorship law. Writing for the court, Judge Mayer held that the scientists at Burroughs Wellcome had already conceived the inventions covering the use of azidothymidine (AZT) to treat HIV before scientists at the National Institutes of Health (NIH) confirmed the drug’s efficacy in laboratory testing. Because conception, not confirmation, fixes inventorship, the NIH scientists were not joint inventors. The opinion’s formulation of the conception standard remains the one courts recite today, including in modern disputes over collaborative and even AI-assisted invention.

At a glance

Case: Burroughs Wellcome Co. v. Barr Laboratories, Inc., 40 F.3d 1223 (Fed. Cir. 1994), Nos. 93-1503, 93-1504, 93-1505
Decided: November 22, 1994 (rehearing denied December 15, 1994); opinion by Judge Mayer, joined by Judge Schall, with Judge Lourie concurring in part and dissenting in part
Holding: Conception is complete when an inventor has a definite and permanent idea of the complete and operative invention; the NIH scientists who confirmed AZT’s activity contributed to reduction to practice, not conception, and were not joint inventors of five of the patents
Status: Final; affirmed in part, vacated in part, and remanded (conception of the ‘750 patent not resolvable as a matter of law on that record)

The doctrinal frame: conception versus reduction to practice

Inventorship is a legal conclusion built on a factual foundation, and the foundational fact is conception. Under 35 U.S.C. § 116, joint inventors are those who contribute to the conception of an invention; a person who merely helps carry out an idea already fully formed by another is not a co-inventor. The stakes are high. Naming the wrong inventors can render a patent unenforceable, and correct inventorship determines ownership, licensing rights, and the ability to sue.

The critical distinction is between conception and reduction to practice. Conception is the formation in the inventor’s mind of the complete idea of the invention. Reduction to practice is either building and testing the invention (actual reduction) or filing a patent application that adequately describes it (constructive reduction). The two can be separated in time. An inventor may conceive an invention long before anyone confirms it works. Burroughs Wellcome is the case that clarified how to tell the two apart, and why the difference decides who is an inventor.

The facts: a drug, a race, and a government laboratory

AZT was originally synthesized in the 1960s as a failed cancer drug. In the mid-1980s, as the AIDS epidemic accelerated, scientists at Burroughs Wellcome screened compounds for activity against retroviruses and formed the idea that AZT might treat what was then called HTLV-III, the virus now known as HIV. Working with murine retrovirus models, the Burroughs Wellcome team developed a specific plan: administer AZT to treat HIV in humans, at particular dosages and via particular methods. They drafted patent applications describing these uses.

Burroughs Wellcome could not test the drug against live HIV in its own facilities, so it sent a coded sample to the NIH, where Dr. Samuel Broder and Dr. Hiroaki Mitsuya ran the assays that confirmed AZT inhibited HIV in human cells. Burroughs Wellcome then filed its patent applications. When Barr Laboratories and Novopharm later sought to market generic AZT, they argued the patents were invalid because Burroughs Wellcome had failed to name the NIH scientists as co-inventors, contending that the invention was not truly conceived until the NIH tests proved it worked.

The holding and the conception standard

The Federal Circuit rejected the challenge as to the patents at issue on appeal. The court articulated the standard that gives the case its lasting authority: conception is “the formation in the mind of the inventor, of a definite and permanent idea of the complete and operative invention.” The idea must be “definite and permanent” in the sense that it involves a specific approach to the particular problem at hand, and it must be sufficiently precise that a person of ordinary skill could reduce it to practice without extensive research or experimentation. Conception must also be corroborated by evidence beyond the inventor’s own testimony.

Applying that standard, the court held that the Burroughs Wellcome inventors had a definite and permanent idea of using AZT to treat HIV, at defined dosages, before the NIH testing. The draft patent applications, prepared before the confirmatory assays, were powerful corroboration that the idea was already complete. What the NIH scientists did, running the assays that confirmed the drug worked, was reduction to practice, not conception. Confirming that a fully conceived invention actually functions does not make the confirming scientist a joint inventor.

The court addressed head-on the objection that the inventors could not have “known” the drug would work before testing. It acknowledged that in some unpredictable fields conception and reduction to practice may be nearly simultaneous, because an inventor cannot be sure of the invention until it is reduced to practice. But that is the exception, not the rule, and it did not apply here. The inventors’ idea was definite and specific before the tests; their lack of absolute certainty that it would succeed went to reduction to practice, which is not part of conception. An inventor need not know that an invention will work to have conceived it.

The court affirmed the inventorship ruling as to five of the six patents (the ‘232, ‘838, ‘130, ‘208, and ‘538 patents). It vacated and remanded as to the sixth, the ‘750 patent, which claimed a method of using AZT to raise the T-lymphocyte count of HIV-infected patients. Because the evidence suggested that a person of ordinary skill might not have expected the T-cell count to rise, the court concluded that “[t]he record does not now support resolution of this question as a matter of law,” that is, whether the inventors had conceived that method before the Phase I clinical trial results. Judge Lourie concurred in the judgment as to the five patents but dissented from the vacatur as to the ‘750 patent. The core conception holding stood.

What it changed

Burroughs Wellcome did not invent the conception doctrine, but it supplied the canonical modern articulation and applied it in a high-stakes, real-world race that made the stakes vivid. The decision drew a firm line: contributing to the confirmation or testing of an already-conceived invention does not confer inventorship. That line protects inventors who rely on outside laboratories, contract researchers, and collaborators to reduce their ideas to practice, and it disciplines the temptation to claim co-inventorship based on hands-on experimental work alone.

The opinion also reinforced the corroboration requirement, insisting that alleged conception be supported by contemporaneous evidence such as notebooks, draft applications, and records. That evidentiary discipline is now standard practice in inventorship disputes and in the way sophisticated research organizations document who thought of what, and when.

Open questions

Burroughs Wellcome set the conception standard but left its application to context. It did not fully resolve how to treat truly unpredictable arts, where the court itself conceded conception and reduction to practice can merge, and later biotechnology cases continue to test that boundary. It addressed the negative question, who is not an inventor, more sharply than the affirmative one, and the companion problem of when a collaborator’s contribution is significant enough to confer joint inventorship was developed in later cases such as Fina Oil & Chemical Co. v. Ewen, 123 F.3d 1466 (Fed. Cir. 1997), and Dana-Farber Cancer Institute, Inc. v. Ono Pharmaceutical Co., 964 F.3d 1365 (Fed. Cir. 2020). And the conception standard’s emphasis on formation of the idea “in the mind of the inventor” now frames a novel debate about inventions generated with the assistance of artificial intelligence, where the locus of the conceiving mind is contested.

Implications for inventors and businesses

Document conception contemporaneously. Dated laboratory notebooks, draft applications, and disclosure records that show a definite and permanent idea are the best proof of who conceived an invention and when.
Distinguish thinkers from testers. A collaborator who confirms that an already-formed idea works is generally not a co-inventor, but one who contributes to the idea itself may be. Get inventorship analysis right before filing.
Manage outside laboratories and CROs carefully. Sending a compound or prototype out for confirmatory testing does not, by itself, create co-inventors, but the underlying agreements should still address intellectual property to avoid disputes.
Take inventorship seriously as a validity risk. Misjoinder or nonjoinder of inventors can be raised to invalidate a patent, so inventorship determinations deserve the same rigor as claim drafting.

Frequently asked questions

What is the legal test for conception under Burroughs Wellcome? Conception is complete when the inventor has formed a definite and permanent idea of the complete and operative invention, specific enough that a person of ordinary skill could reduce it to practice without extensive research or experimentation. The idea, not the later confirmation that it works, marks conception.

Why weren’t the NIH scientists joint inventors of the AZT patents? The Burroughs Wellcome inventors had already conceived the idea of using AZT to treat HIV and had written it into draft patent applications before the NIH tests. The NIH scientists confirmed the drug worked, which is reduction to practice, not conception. Confirming that a conceived idea works does not make one a co-inventor.

Does an inventor have to know the invention will work to have conceived it? Generally no. The court recognized that in unpredictable fields conception and reduction to practice can nearly merge, but on these facts the inventors’ idea was definite and permanent before testing. Certainty that the invention will work is reduction to practice, and it is not required for conception.