Week 12 - Defenses to Patent Infringement (III)

Walker Process and Handgards Claim

Two Ways a Patent Owner Can Break Antitrust Law A patent owner can face antitrust liability for enforcing a patent it got by fraud, which is called a Walker Process claim. It can also be liable for enforcing a patent it knows is invalid or not infringed just to hurt a competitor, which is called a Handgards claim.

Nobelpharma AB v. Implant Innovations, Inc. 141 F.3d 1059 (Fed. Cir. 1998) Background

The Dental Implant Patent The patent covered titanium dental implants with tiny "micropits" that help the implant bond to the jawbone. One inventor, Branemark, had co-written a 1977 book showing images of implants with micropits that appeared to fall within the patent's size range.

The Book Was Left Out An inventor's draft patent description mentioned the 1977 book, but the patent agent removed every reference to it. The book was also never mentioned in the U.S. patent application.

Procedural Posture

The Nobelpharma Appeal Nobelpharma (NP) appealed after losing a jury verdict on an antitrust counterclaim brought by Implant Innovations (3I). The Federal Circuit upheld the trial court's refusal to overturn that verdict or order a new trial.

NP Sues 3I NP licensed the technology in 1980, the patent issued in 1982, and NP sued 3I for infringement in 1991. 3I fought back with an antitrust counterclaim, saying NP sued knowing the patent was invalid because the book had been hidden from the Patent Office (PTO).

The Trial Court Rules the Patent Invalid Branemark admitted in an old deposition that key manufacturing details might have been kept as trade secrets and left out of the patent. The court ruled the patent invalid for failing to disclose the best way to make the invention and let the antitrust claim go to the jury.

The Jury Finds Fraud The jury found that the patent was obtained by fraud and that NP knew this when it sued to block 3I from competing. It awarded about $3.3 million, which the court tripled under the antitrust laws.

The Trial Court Upholds the Verdict The trial court refused to overturn the verdict because the jury found knowing fraud on the PTO and enforcement with knowledge of that fraud. It also ruled that the jury did not need to decide whether NP's lawsuit was "objectively baseless".

The Court Sides with 3I The Federal Circuit agreed that strong evidence showed the patent was obtained by fraud and enforced with knowledge of that fraud. It also found that NP was not harmed by any error in the jury instructions.

Legal Arguments

NP's Arguments on Appeal NP argued that the evidence did not show fraud or that it knew of any fraud, and that such findings cannot support antitrust liability anyway. It also claimed it deserved a new trial because the jury was never told that a lawsuit must be "objectively baseless" to create liability.

3I's Response 3I said strong evidence supported the fraud findings and that no "objectively baseless" instruction was needed because actual knowledge of fraud had to be proven. It added that NP had waived the point by never asking for that instruction.

Legal Standard

Federal Circuit Law Will Govern Because these issues arise in patent cases that the Federal Circuit hears, the court decided to apply its own law rather than the law of the various regional circuits. It overruled earlier cases that said otherwise but will still use regional circuit law for general antitrust issues such as market definition and damages.

Two Routes to Losing Antitrust Protection A patent owner who sues can lose antitrust protection if the patent was obtained by knowing, willful fraud, or if the lawsuit was a "sham" meant to interfere with a competitor's business. Under Walker Process, the challenger must prove the patent owner knowingly lied to the PTO.

Knowledge at the Time of Suit The patent owner must also have known about the fraud when it filed suit. The Supreme Court based this rule on earlier cases involving deliberate lies, such as false invention dates and fake expert praise.

Justice Harlan's Warning Justice Harlan said courts must separate deliberate fraud from patents that fail for technical reasons. Otherwise, the fear of triple-damage lawsuits could discourage inventors from seeking patents and sharing their inventions.

Inequitable Conduct Is a Lesser Offense The court distinguished "inequitable conduct" from Walker Process fraud, noting that inequitable conduct is broader and covers less serious misconduct. Its predecessor court defined true fraud as conduct so serious that it could support a lawsuit on its own.

The Five Elements of Fraud Fraud requires a false statement of an important fact, intent to deceive, reasonable reliance by the deceived party, and resulting harm. Lesser misconduct that misses one of these elements is judged by looking at the overall fairness of the case.

A Shield Versus a Sword Inequitable conduct is a defense that can make a patent unenforceable and lead to attorney fees. Fraud is more serious because it can be used offensively to impose antitrust liability, including triple damages.

Omissions Can Be Fraud The court held that hiding information can be just as bad as lying, so a fraudulent omission can support antitrust liability. The omission must show a clear intent to trick the examiner into granting a patent that should not have issued.

A Higher Bar for Fraud Walker Process fraud requires stronger proof of intent and importance than inequitable conduct, with no trading off a weak showing on one against a strong showing on the other. Simply failing to cite a reference is not enough; there must be proof of fraudulent intent and that the patent would not have issued otherwise.

The Trial Court's Approach to the Sham Test The trial court noted that the Supreme Court had left open how the "sham" test applies to patent applications. After reviewing other circuits, it decided the sham test does not apply to fraud-based claims, but the Federal Circuit disagreed with that reasoning.

Two Separate Theories The Federal Circuit held that Walker Process fraud and sham litigation are two independent ways to lose antitrust protection, and either or both can apply. A fraud claim does not require the extra sham inquiry because knowingly enforcing a fraudulent patent is clearly wrongful on its own.

The Two-Part Sham Test To show a sham, a challenger must prove that no reasonable litigant could expect to win and that the suit was really meant to harm a competitor through the process itself. Even then, the challenger must still prove every other element of an antitrust violation.

Sham Focuses on the Lawsuit If a suit is not objectively baseless, the patent owner's motives do not matter. Unlike a fraud claim, a sham claim looks at the lawsuit itself, not how the patent was obtained.

Court Reasoning

Evidence of Intent to Deceive The court found strong evidence that the 1977 book was deliberately hidden from the PTO. The jury could reject the patent agent's claim that he simply did not think the book was relevant, and could infer that Branemark, as the book's author, knew it mattered.

The Book Mattered An expert testified that the book's images showed implants with every element of the patent claims, including the claimed micropit sizes. Even Branemark admitted that the micropit sizes were easy to calculate, so the jury could find the patent would not have issued if the PTO had seen the book.

NP Knew of the Fraud An NP officer testified that a colleague said hiding the book would be "fraud". NP's own lawyer had warned that any lawsuit on the patent would fail because the book would invalidate it.

The Jury Instructions Were Proper The court found that the fraud instructions required clear and convincing evidence, knowing and intentional acts, and proof that the patent would not have issued otherwise. Because the fraud finding was sound, NP gained nothing from extra instructions on the sham test, and its other objections failed.

Conclusion of Nobelpharma The court found that the trial court made no errors in its rulings on evidence or in denying a new trial. A reasonable jury applying the correct law could find fraud under Walker Process.

Settlements Raise Issues

Settlements Can Raise Antitrust Concerns Courts generally favor settlements because patent litigation is costly and settlements reflect what the parties want. But settlements between competitors can harm competition, especially "reverse payments," in which the patent owner pays the accused infringer to stay out of the market.

Federal Trade Commission v. Actavis, Inc. 133 S. Ct. 2223 (2013) Background

The Actavis Question In Actavis, the patent owner paid the alleged infringer millions to delay selling its product. The Supreme Court asked whether such "reverse payment" deals can unreasonably reduce competition in violation of antitrust law.

The Drug Approval Framework Most reverse payment deals arise under the Hatch-Waxman Act, which governs generic drug challenges to brand-name drug patents. Brand-name drug makers must go through long, costly FDA testing to win approval for a new drug.

Generics Get a Shortcut Generic makers can use an abbreviated application showing that their drug is equivalent to an approved brand-name drug, avoiding costly studies. This speeds cheaper generics to market and increases competition.

Patent Disputes Under Hatch-Waxman Brand-name makers must list their relevant patents, and generics must certify how they will avoid infringing them, including by claiming the patent is invalid or not infringed (a "paragraph IV" challenge). A paragraph IV filing usually triggers a lawsuit, which delays FDA approval of the generic for up to 30 months.

The 180-Day Reward The first generic to file a paragraph IV challenge gets 180 days during which no other generic can compete, a period that can be worth hundreds of millions of dollars. Only the first filer can get this period, and no other generic can claim it if the first filer loses it.

The AndroGel Dispute Solvay held a patent on the brand-name drug AndroGel, and generic makers Actavis and Paddock challenged it, with Par sharing Paddock's costs. Solvay sued, and in 2006 all the parties settled.

The Settlement Terms The generic makers agreed to stay off the market until 2015, and Solvay agreed to pay them millions, including $19–30 million a year to Actavis. The FTC said these payments, labeled as fees for other services, were really payments to delay competition.

Procedural Posture

The Lower Court Was Wrong The Eleventh Circuit had dismissed the Federal Trade Commission's (FTC's) case, holding that such deals are immune if their effects stay within the patent's scope. The Supreme Court disagreed, holding that these deals can sometimes violate antitrust law and that the FTC's suit should go forward.

The FTC Sues and Loses at First The FTC sued in 2009, claiming the companies agreed to share Solvay's monopoly profits and keep cheap generics off the market. The district court dismissed the case, finding no antitrust violation.

The Eleventh Circuit's Reasoning The Eleventh Circuit held that, without sham litigation or fraud, such deals are immune if their effects stay within the patent's scope. It reasoned that patent owners have a legal right to exclude others and that courts should not force parties to keep litigating.

The Supreme Court Takes the Case The Supreme Court agreed to hear the case because lower courts had reached conflicting results on these settlements. It accepted that the deal's effects might fall within the patent's potential to exclude, but held that this does not make the deal immune.

Court Reasoning

The Patent Might Be Invalid A valid patent lets its owner exclude others, but an invalid patent gives no such right, and even a valid patent cannot block products that do not infringe it. The settlement ended a lawsuit over exactly those questions, and paying a defendant to stay out of the market is unusual and worrying.

Both Patent and Antitrust Policy Matter The Court said it would be odd to judge these deals only by patent policy and ignore antitrust policy. Its past cases weighed traditional antitrust factors, such as harm to competition, benefits, and market power, together with patent considerations.

The Dissent's Views Are the Novel Ones The majority said its approach is not new and that the dissent's idea of paying a rival to drop a patent challenge without any scrutiny is what is truly novel. No patent law gives that right, and it conflicts with the policy of eliminating unjustified patents.

Ordinary Settlements Are Different It is common for a defendant to pay less than the full damages claimed, or for a plaintiff to pay to settle a counterclaim, and the Court did not question those deals. But in a reverse payment, a party with no damages claim is paid simply to stay out of the market, which is quite different.

Settlement Policy Does Not Decide the Case The Eleventh Circuit worried that antitrust review would require costly litigation over patent validity. The Court recognized the value of settlements but gave five reasons the FTC should be allowed to prove its case.

Reason 1: Real Harm to Competition A reverse payment lets the patent owner buy protection from losing its patent and keep monopoly prices, sharing the profits with the challenger while consumers lose. A payment larger than what the generic could earn by winning suggests that the goal is to split monopoly profits.

Why Other Challengers May Not Step In Hatch-Waxman's 180-day reward goes only to the first challenger, so later challengers have less reason to sue. Later challengers also face the same 30-month delay, so paying off the first filer removes the most motivated competitor.

This Mostly Happens in Pharma Scholars note that, outside the drug industry, it is almost unheard of for a patent owner to pay an accused infringer to settle. Hatch-Waxman's unique rules likely explain why these deals happen more often with drugs.

Reason 2: Often Unjustified Some reverse payments may be justified, for example as saved litigation costs or fair pay for real services. But defendants can prove those justifications at trial, so they are no reason to dismiss the FTC's case.

Reason 3: Market Power A large payment to keep a rival out is itself a strong sign that the patent owner has power to charge above-competitive prices. Firms without that power are unlikely to pay large sums for others to stay away, and studies link these deals to higher profits.

Reason 4: Courts Can Manage These Cases Courts usually do not need to decide whether the patent is valid to resolve the antitrust question. A large, unexplained payment suggests that the patent owner doubts its patent and wants to share monopoly profits rather than face competition.

Payment Size as a Stand-In Even if a valuable patent's owner might pay a lot to avoid a small risk, an unexplained payment likely aims to prevent competition. So the payment's size can stand in for the patent's weakness without a full validity trial.

Reason 5: Parties Can Still Settle Parties can still settle in other ways, such as letting the generic enter before the patent expires without any payment. If the reason for a payment is to keep and share monopoly profits, antitrust law likely forbids it.

Summary of the Five Reasons Large, unjustified reverse payments can harm competition, may reflect patent-based market power, and can be judged by their size without a validity trial. Together, these points outweigh the policy favoring settlement that led the Eleventh Circuit to grant near-automatic immunity.

No Presumption of Illegality The FTC asked the Court to treat reverse payments as presumed illegal under a quick review. The Court refused because such shortcuts apply only when the harm is obvious, and that is not true here.

Full Rule-of-Reason Review Whether a payment harms competition depends on its size, how it compares with litigation costs, whether it pays for real services, and other justifications. So the FTC must prove its case under the standard "rule of reason," but it need not litigate patent validity or every possible theory.

Dissent - discouraged settlements

The Dissent's Core View Chief Justice Roberts, joined by Justices Scalia and Thomas, said the only question should be whether the deal gave Solvay more power than its patent already did. He argued that the majority's approach has no basis in any law and will discourage settlements.

Patents Are an Exception to Antitrust Law Antitrust law promotes competition, while patent law grants limited monopolies to encourage invention. A patent creates a zone where its owner can act without antitrust liability, and settling patent suits has never been illegal in itself.

Settlements Within the Patent's Scope Settlements within the patent's scope are not subject to antitrust review, except for settlements of sham suits or suits on patents obtained by fraud. Neither exception applied here, so Solvay simply paid competitors to respect its patent, which the dissent saw as a good thing.

The Majority Will Discourage Settlement Parties will not settle if they must then relitigate patent validity in an antitrust case. A patent is either valid or not, and uncertainty about the answer should not create liability.

Large Payments Do Not Prove Doubt A patent owner may be 95% confident in its patent but still pay a lot to avoid even a 5% risk. Figuring out its true motives will be hard, especially because legal advice is usually protected from disclosure.

Absurd Results Under the majority's view, a first settlement could be illegal even if the patent is later held valid by the Supreme Court. The patent owner would be punished just for settling instead of litigating to the end.

The Logic Has No Limit The idea that removing any chance of invalidation is an antitrust problem could reach licenses and even the earlier-entry deals the majority recommends. District courts will struggle to apply the vague rule-of-reason factors.

It May Discourage Generic Challenges Ironically, limiting settlements lowers the expected payoff for generics, which may make them less willing to challenge drug patents at all. Parties settle more easily when they have more valuable things to trade.

Dissent's Conclusion The dissent said the majority weakens patent protections, undercuts the policy favoring settlement, and may defeat its own goal. It would have kept patent questions out of open-ended antitrust review with its triple damages and heavy discovery.

Refusal to Deal

The Right Not to License A patent owner has the right to exclude others, including the right not to use, sell, or license the patent. That right has limits, though.

In re Independent Service Organizations Antitrust Litigation 203 F.3d 1322 (Fed. Cir. 2000) Background

The Xerox Case CSU sued Xerox for refusing to sell it patented parts and to license copyrighted software and manuals. The Federal Circuit upheld summary judgment for Xerox.

Xerox's Parts Policy Beginning in 1984, Xerox refused to sell unique copier parts to independent service companies unless they were end users, and it later tightened enforcement. CSU kept its business going with parts taken from used machines, parts from other companies, and parts bought through customers.

Procedural Posture

The Lawsuit Xerox stopped its European affiliate from selling parts to CSU, and CSU opted out of a 1994 class settlement to sue on its own. CSU claimed that Xerox overpriced parts to drive service companies out of business, and Xerox counterclaimed for infringement.

The District Court's Ruling The district court held that lawfully obtained patents and copyrights let their owners refuse to sell or license, even if competition in other markets is affected. It ruled that Xerox's motives were irrelevant.

Legal Standard

Which Law Applies The Federal Circuit applied its own law to the patent issues and the Tenth Circuit's law to the copyright issues. This followed Nobelpharma's rule that patent-related antitrust immunity is a Federal Circuit question.

Cour Reasoning

Patents Do Not License Antitrust Violations Intellectual property does not permit antitrust violations, but antitrust law does not cancel a patent owner's right to exclude. A patent alone does not prove market power, and even real market power does not create a duty to license.

No Case Has Punished a Refusal No reported case had imposed antitrust liability for a one-sided refusal to sell or license a patent. The Patent Act also says that refusing to license is not patent misuse.

The Limits on the Right to Exclude A patent owner who sues is exempt from antitrust law unless the patent was obtained by fraud or the suit is a sham. CSU did not claim fraud, so Walker Process did not apply.

No Sham Claim CSU alleged that Xerox misused its patents but did not claim that Xerox's infringement counterclaims were shams. If a suit is not objectively baseless, motive is irrelevant.

The Kodak Footnote CSU relied on Kodak to argue that Xerox used its patent power to dominate the service market. The court said Kodak was a tying case with no patents asserted, and that it only bars using a patent to gain a monopoly beyond the patent's scope.

Rejecting the Ninth Circuit's Approach The Ninth Circuit had presumed that refusals to license are justified but allowed juries to find that justification was a pretext. The Federal Circuit declined to follow that approach because it would require probing the patent owner's motives.

No Inquiry into Motive Without illegal tying, fraud, or sham litigation, a patent owner may enforce its right to exclude free from antitrust liability. Courts will not examine its motives so long as it does not extend its power beyond the patent.

Xerox Wins The accused infringer bears the burden of proving an exception, and CSU did not do so. Xerox had no duty to sell or license its patented parts and did not violate antitrust law.

The Duty of Candor

Patent applicants and their lawyers must disclose information that is important to patentability to the PTO. Violating this duty is inequitable conduct, the subject of Therasense.

Therasense, Inc. v. Becton, Dickinson & Co. 649 F.3d 1276 (Fed. Cir. 2011) (en banc) Background

The Test Strip Patent The strips use an electrochemical sensor to measure glucose in a drop of blood. The patent claims a strip that tests whole blood without a protective membrane over the electrode.

Why Membranes Mattered Earlier sensors used membranes to control glucose flow or to stop red blood cells from "fouling" the electrode. Abbott filed its application in 1984, and it was repeatedly rejected over 13 years.

Abbott's New Argument Abbott's earlier '382 patent said a membrane was "optionally, but preferably" used with live blood. In 1997, Abbott argued to the PTO that experts would have read this to mean a membrane was required for whole blood.

The Statements to the PTO Abbott's research director declared that experts would not read the language as making a membrane optional. Abbott's lawyer added that the phrase was mere "patent phraseology," not a technical teaching.

Contradictory Statements in Europe Years earlier, Abbott told the European Patent Office that the membrane in the same patent was optional. In a second brief, it called this disclosure "unequivocally clear" that the membrane is optional.

Procedural Posture

The Therasense Case The district court held Abbott's blood glucose test strip patent unenforceable for inequitable conduct. The full Federal Circuit vacated that ruling and sent the case back.

The Litigation Becton sued Abbott in 2004, and Abbott countersued for infringement of three patents. The district court held the '551 patent unenforceable because Abbott never gave its European briefs to the PTO.

Rehearing to Fix the Doctrine The court reheard the case as a full court because inequitable conduct had been overused. Inequitable conduct grew out of a doctrine called "unclean hands" and, if proved, bars enforcement of the patent.

Legal Standard

How the Doctrine Expanded The original cases involved extreme misconduct like perjury and fake evidence, but the doctrine grew to cover mere nondisclosure. It also adopted a harsher remedy, making the whole patent unenforceable, and required proof of both intent and importance by clear and convincing evidence.

Unintended Consequences Lowered standards made inequitable conduct a common litigation tactic that expands discovery and disqualifies the patent lawyer from the case. It also smears patent owners, discourages settlement, and adds cost and complexity.

The "Atomic Bomb" Inequitable conduct makes the entire patent unenforceable, cannot be fixed, and can spread to related patents. It can also trigger antitrust claims and fee awards, so the court tightened the standards for both intent and importance.

Specific Intent Required The challenger must prove a specific intent to deceive the PTO; negligence or "should have known" is not enough. It must show the applicant knew of the reference, knew it was important, and deliberately withheld it.

No Sliding Scale Intent and importance must be proven separately, and intent cannot be inferred from importance alone. Intent may be inferred from circumstances, but it must be the single most reasonable inference.

Multiple Inferences Defeat Intent If more than one reasonable inference is possible, intent cannot be found. The patent owner need not explain itself until intent is first proven, and failing to give an explanation does not prove intent.

"But-For" Materiality Information is important ("material") only if the PTO would not have granted the claim had it known about it. Courts judge this under the PTO's easier standard, so a reference can be material even if it does not invalidate the patent in court.

The Remedy Should Fit the Wrong Because unenforceability is severe, it should apply only when the misconduct actually produced an unwarranted patent claim. Enforcing a valid patent does not harm the public just because of unrelated misconduct.

The Egregious Misconduct Exception An exception applies to affirmative egregious acts, like filing a clearly false affidavit, which count as material automatically. A patent owner would not go to such lengths to deceive unless it believed the lie would matter.

Omissions Need But-For Proof Simply failing to disclose a reference is not egregious misconduct, so it requires but-for proof. This balances encouraging honesty against unfounded accusations, while the exception covers extraordinary cases.

Court Reasoning

Rejecting the PTO's Rule 56 The court declined to adopt the PTO's definition of materiality in Rule 56. Tying the doctrine to PTO rules, which change over time, caused uncertainty and the very problems the court wanted to fix.

Rule 56 Is Too Broad Rule 56 treats information as material even if the patent owner's explanation would make it irrelevant, and covers anything marginally related to patentability. Adopting it would keep applicants flooding the PTO with marginal references and litigators charging inequitable conduct in nearly every case.

Applying the New Test The district court had used Rule 56 rather than the but-for test, so its materiality findings were vacated. On remand, it must decide whether the PTO would have rejected Abbott's arguments if it had seen the European briefs.

Intent Findings Vacated The district court wrongly based intent on Abbott's lack of a good-faith explanation and on a "should have known" standard. On remand, it must decide whether clear evidence shows that the two individuals knowingly and deliberately withheld the briefs to deceive the PTO.

Experimental Use

Two Kinds of Experimental Use The statute exempts uses reasonably related to developing information for FDA approval, known as the Bolar or FDA exemption. The judge-made exemption is weak, and the Federal Circuit has never used it to excuse infringement.

Merck v. Integra Lifesciences I 545 U.S. 193 (2005) Background

The Merck Question The Supreme Court asked whether using patented inventions in early lab research that is never submitted to the FDA is exempt from infringement. Using a patented invention is normally infringement, but Congress created this exemption in 1984.

The Exemption's Text The law exempts uses solely for purposes reasonably related to developing and submitting information under federal drug laws. Drug makers submit data twice: to get permission for human trials and to get approval to sell the drug.

The RGD Peptide Patents Integra owned patents on the "RGD peptide," which helps cells stick together. Merck funded Dr. Cheresh at Scripps, who found that blocking certain cell receptors could stop new blood vessel growth and reverse tumors.

The Research Deal In 1995, Merck pledged $6 million over three years for research aimed at an application to begin human trials. Scripps tested Merck's peptides for effectiveness and toxicity and picked the best candidate in 1997.

Other Research and FDA Steps Scripps also did basic research using the peptides as comparison controls. Merck began the approval process, discussed the drug with the FDA, and got the National Cancer Institute to sponsor trials.

Procedural Posture

The Infringement Suit Integra sued Merck, Scripps, and Dr. Cheresh in 1996. Merck argued that it did not infringe and that its research was protected by the research exemptions.

The Trial and Appeal The jury found infringement and awarded $15 million. The Federal Circuit affirmed, calling the work general research to find new drugs rather than testing to supply the FDA.

The Lower Court's Reasoning The Federal Circuit held that the research only identified the best drug candidate and that the FDA has no interest in the hunt for drugs. It also held that the exemption does not cover all research that might someday lead to FDA approval.

Court Reasoning

A Wide Safe Harbor The Supreme Court said the exemption gives wide room for using patented drugs in activities tied to FDA regulation. It covers all uses reasonably related to submitting any information to the FDA, including early lab studies.

The FDA Wants More Than Safety Data Integra argued that the FDA cares only about early safety data. The Court disagreed, noting that the FDA requires summaries of a drug's effects, toxicity, and behavior in animals.

Risk-Benefit Review The FDA weighs risks against benefits when approving trials, so evidence of a drug's effectiveness matters too. Early studies on effectiveness are therefore relevant to the FDA.

Lab Practice Rules Do Not Disqualify The FDA's "good laboratory practices" rules apply only to safety studies, not to studies on effectiveness or how a drug works. Even safety studies that do not follow those rules can be submitted with an explanation.

Basic Research Is Not Covered The Supreme Court agreed that basic research without a goal of developing a particular drug is not covered. But the exemption can still cover drugs that are never submitted and experiments that are never submitted.

Science Is Trial and Error Researchers cannot know in advance which promising candidates will succeed. Limiting the exemption to drugs that are ultimately submitted would effectively restrict it to generics, which Congress did not intend.

The Reasonable Basis Test Use is protected if the drug maker reasonably believes that the compound may produce a particular effect and the research, if successful, would be fit to submit to the FDA. Leaving an experiment's data out of the final submission does not make the use infringing.

Uncertainty About What to Submit Companies often do not know exactly what information the FDA will require, especially at early stages. So early studies are protected if there is a reasonable basis to believe that they will produce relevant information.

Case Sent Back The Federal Circuit used the wrong reading of the law in reviewing the evidence. The Supreme Court vacated the judgment and sent the case back.

Madey v. Duke University 307 F.3d 1351 (Fed. Cir. 2002) Background

Madey Moves to Duke Madey, a highly regarded Stanford laser researcher, moved his lab to Duke in 1989. He owned two patents used by the lab's equipment and ran the lab for almost a decade.

The Dispute Duke said Madey managed the lab poorly, while Madey said Duke misused the lab's funding. Duke removed him in 1997, he resigned in 1998, and Duke kept using the equipment.

The Patents and Devices The patents covered a microwave electron gun and a laser oscillator. Duke seemed to concede that its three devices used the patented technology.

Procedural Posture

The Madey v. Duke Case Dr. Madey sued Duke University for infringing his laser patents. The Federal Circuit held that the district court was wrong to apply the experimental use defense.

The District Court's Defense The district court recognized an exception for uses solely for research, academic, or experimental purposes. It relied on a case warning that the defense cannot cover inquiry with real commercial purposes.

The District Court Rejects It The court relied on Duke's policy stating that it does not do research mainly to develop patents. It found Madey's evidence speculative and ruled that he had not proven infringement.

Legal Arguments

Burden on Madey The district court required Madey to prove that Duke's use was not solely experimental or nonprofit. Madey argued that this wrongly shifted the burden to him.

Madey's Commercial Argument Madey argued that Duke's research was commercial, pointing to a case finding that government helicopter testing served its business. He said Duke's patent policy showed that it seeks grants and commercial uses.

Madey's Three Claimed Errors Madey argued that the court shifted the burden, used too broad a defense, and relied on weak evidence. The Federal Circuit largely agreed on all three points.

Court Reasoning

The Defense Still Exists, Barely Madey argued that a Supreme Court case abolished the experimental use defense. The court held that the defense survives, but only in a very narrow form.

Not an Affirmative Defense Madey argued that Duke had to raise the defense in its first response or lose it. The court disagreed, finding no authority for that rule and allowing the defense to be raised later.

The Burden Was Wrongly Shifted The district court made Madey prove that the use was not experimental as part of his basic infringement claim. That was wrong because Duke, not Madey, must establish the defense.

The Defense Is Very Narrow The district court's version, covering research, academic, or nonprofit uses, was far too broad. The defense covers only uses for amusement, idle curiosity, or strictly philosophical inquiry, and not uses with any commercial purpose.

Legitimate Business Disqualifies Use that furthers the user's legitimate business does not qualify, even with no commercial purpose. The district court relied on an old case that wrongly treated a school's educational use as automatically exempt.

Universities Have Business Too University research furthers legitimate business goals, such as educating students, raising prestige, and attracting grants and faculty. So any use that furthers the user's business and is not purely for curiosity is not protected.

Nonprofit Status Does Not Decide It Whether the user is a nonprofit is not decisive, and the district court gave Duke's status too much weight. On remand, the court must focus on Duke's legitimate business and whether the use was purely for curiosity.