Abstract
Congress enacted the Hatch-Waxman Act (“Hatch-Waxman”) in 1984 to facilitate and hasten market entry of generic versions of brand drugs, in order to increase competition and bring down the cost of drugs. 1 One of the key features of Hatch-Waxman is the ability of a branded drug manufacturer to bring a cause of action for patent infringement against a would-be generic competitor under 35 U.S.C. § 271(e)(2), based on the contents of the generic's Abbreviated New Drug Application (ANDA). These lawsuits differ fundamentally from traditional patent infringement litigation in that they are filed prior to a product actually entering the market. When the provisions of Hatch-Waxman operate as intended, infringement litigation will be resolved prior to the generic drug entering the market. Hatch-Waxman provides for a mandatory 30-month stay of ANDA approval when the branded drug company sues the ANDA applicant under Hatch-Waxman, presumably in order to allow enough time for the parties to resolve the dispute.
Hatch-Waxman litigation thus generally occurs prior to any actual infringement, because the generic is not yet on the market. 2 This feature of Hatch-Waxman litigation represents a marked departure from other patent infringement litigation, which generally occurs after some act of alleged infringement. 3 In Hatch-Waxman litigation, the question for the court is not whether a valid patent has been infringed, but whether the proposed generic product would infringe a valid patent if it enters the market prior to expiration of the patent. The Federal Circuit often refers to infringement alleged under § 271(e)(2) as an “artificial” act of infringement. 4
There are a number of desirable consequences that flow from allowing a branded drug company and ANDA applicant to resolve a patent dispute under § 271(e)(2) prior to generic market entry. For example, it eliminates any possibility of an award of monetary damages in cases in which the ANDA application is found to infringe. Hatch-Waxman specifically provides that the only remedy available to the branded drug company is an injunction delaying approval of the ANDA until after the infringed patent has expired. 5 This is significant, because the amount of damages that might be awarded for the actual sale of an infringing generic product can vastly exceed the profit a generic company might make off the sale of the product. This is particularly the case if the amount of damages is calculated based on the branded drug company's “lost profit.” An example of this can be seen in GlaxoSmithKline v. Teva, a recent Federal Circuit decision that is the focus of this article. 6
Due to the potential for a huge award of damages in a case in which a generic drug enters the market and then was later found to be infringing, the term “at-risk” is often used to describe market entry by a generic product prior to resolution of patent infringement litigation with the branded drug manufacturer. An “at-risk” generic product launch is risky for both sides: obviously the generic company risks a judgment vastly exceeding revenue generated by the sale of the generic product, but the branded drug company also faces the risk that a generic company, particularly a small underfunded one, might be unable to pay the judgment, leaving the branded company uncompensated for its loss. Hatch-Waxman's 30-month stay of ANDA approval after the filing of a § 271(e)(2) lawsuit is presumably intended to discourage at-risk generic launches, by giving the parties two and half years to resolve the dispute prior to the generic being permitted to enter the market. In some cases 30 months is not enough, and the ANDA might be approved before resolution of the litigation, giving the generic company the option of bringing the product to market. Generic companies will often avoid the potential huge liability of an at-risk launch by waiting for a resolution to the infringement litigation, but sometimes generic companies decide to go ahead and launch at-risk in spite of the potentially massive liability.
Branded drug companies employ various types of patents to maintain exclusivity of their products. The core patent is typically directed towards the active ingredient itself as a composition of matter, but other patents often serve to extend the period of exclusivity beyond the term of the composition of matter patent. Some of these patents are directed towards aspects of the pharmaceutical product other than the active ingredient, including patents on the drug's formulation. In the context of ANDA litigation, patents directed towards a product themselves will be found infringed by an ANDA that proposes marketing a generic falling within the scope of the patent. The sale of the generic product in the U.S. will directly infringe the product patent, because selling (or offering to sell) a product covered by a patent constitutes infringement. 7
In addition to product patents, branded drug companies often obtain patents on methods of using the drug to treat a specific medical condition, which I will refer to this article as “method of treatment” patents. For example, in GlaxoSmithKline, the patent at issue is a method of treatment patent claiming methods of using the branded company's drug for the treatment of heart failure. Significantly, the sale of a generic drug does not constitute direct infringement of a method of treatment patent, for the simple fact that the generic company is not itself using the drug to treat the disease. Instead, the only potential direct infringers of the patent are the patients who use the drug to treat the medical condition, and the doctors who prescribe the drug for that purpose.
A generic drug company can, however, be found liable for infringement of a method of treatment patent under a theory of indirect patent infringement. There are two forms of indirect infringement, inducement to infringe and contributory infringement. Both are explicitly provided for in the Patent Act under Sections 271(b) and 271(c), respectively. 8 With respect to generic drugs, inducement to infringe plays a much more significant role than contributory infringement, for reasons discussed below. In either case, the generic company is held liable for inducing (or contributing to) the infringement of the direct infringers, i.e., doctors and their patients.
To establish liability under § 271(c) for contributory infringement, a patentee must prove that an accused contributory infringer sold a product having no substantial noninfringing use, i.e., a non-staple product. In the context of method of treatment patents, this means that there generally will be no viable cause of action for contributory infringement if the drug in question has a substantial noninfringing use. There are cases in which a drug only has one substantial use, and there can be liability for contributory infringement if that use is covered by a method of treatment patent. This happened, for example, in Eli Lilly & Co. v. Actavis Elizabeth LLC, where the patent at issued claimed a method of using atomoxetine to treat attention deficit hyperactivity disorder (ADHD). 9 The Food and Drug Administration (FDA) had only authorized use of atomoxetine for one indication, treatment of ADHD. The accused infringer argued that atomoxetine had substantial noninfringing off-label use, amounting to as much as 20 percent of the total use. The patent owner conceded that up to eight percent of the total use of the drug was off-label. However, the generic product sold by the defendants was specifically labeled only for use in the treatment of ADHD, and the Federal Circuit pointed out that generic drug companies are restricted from selling a federally regulated drug for unapproved uses. 10 The court rejected the defendants' argument that physicians are permitted to prescribe atomoxetine for unauthorized use and held that “such unauthorized activity does not avoid [contributory] infringement by a product that is authorized to be sold solely for the infringing use.”
Notably, in Eli Lilly & Co. v. Actavis Elizabeth LLC, the Federal Circuit also found the generic drug company defendants liable for inducement to infringe the patent, based on the product's labeling, which provided instructions for use of the drug in the treatment of ADHD. As a general matter, the marketing of a generic version of a drug that leads to liability for contributory infringement of a method of treatment patent will almost inevitably lead to liability for induced infringement. This is because of FDA labeling requirements, discussed in more detail below, which effectively require a generic company selling a drug with only a single approved use to include that use on its label. It is this labeling that results in liability for induced infringement.
In fact, I am unaware of any Federal Circuit decision which found a generic drug company liable for contributory infringement of a method of treatment patent, but not liable for inducement of infringement. The closest I know of is a case involving an external defibrillator, a type of medical device. The case is Koninklijke Philips N.V. v. Zoll Med. Corp., in which the Federal Circuit held as a matter of law that the sale of an unauthorized defibrillator constituted contributory infringement of a patent relating to a “self-test” function of the device. 11 The court did not address inducement to infringe, presumably considering the issue moot in light of the finding of contributory infringement.
Artificial Inducement Under Hatch-Waxman
Section 271(b) of the Patent Statute provides that “[w]hoever actively induces infringement of a patent shall be liable as an infringer.” 12 Supreme Court and Federal Circuit precedent has established that liability for inducement of infringement requires a showing that the accused took active steps to aid and abet the direct infringement of another, such as by promoting, advertising, or encouraging others to directly infringe, and that the accused knew that the induced conduct is likely to infringe a patent. Thus, for example, a party unaware of the existence of a patent generally cannot be held liable for inducement of infringement. 13 The knowledge element of induced infringement is generally not an issue with respect to generic drugs; ANDA applicants are generally well aware of the patents at issue. When Hatch-Waxman is functioning as designed, product patents and method of treatment patents relating to the product are listed in the Orange Book. But GlaxoSmithKline provides an example of litigation involving allegations of infringement of a patent that was not listed in the Orange Book (and in fact had not even issued) at the time the generic drug entered the market. The deciding issue in most ANDA litigation involving allegations of induced infringement of a method of treatment patent is whether the generic company took active steps to promote the infringing use of the drug.
In the vast majority of cases involving generic drugs, the necessary element of instruction, promotion, and/or encouragement to infringe required to support a finding of inducement under § 271(b) resides in the FDA-approved labeling for the generic product. In 2011 the Federal Circuit noted that “[w]e have long held that the sale of a product specifically labeled for use in a patented method constitutes inducement to infringe that patent, and usually is also contributory infringement.” 14 In cases where an ANDA applicant's proposed label directly instructs use of the drug in a manner that infringes the method of treatment patent, the Federal Circuit has on a number of occasions found the generic company liable for induced infringement. 15
On the other hand, generic companies have been found not to be liable for inducement of infringement in cases in which the label does not describe an infringing use of the drug. In Warner-Lambert v. Apotex Corp., for example, the court considered whether an ANDA applicant infringed a patent claiming an unapproved method of treatment, i.e., an off-label use. 16 Naturally, the proposed label for the generic product did not include an indication for that method. The court explained that “Congress recognized that a single drug could have more than one indication and yet that the ANDA applicant could seek approval for less than all of those indications,” and held that the generic label did not constitute an act of inducement under § 271(b), explaining that “the request to make and sell a drug labeled with a permissible (noninfringing) use cannot reasonably be interpreted as an act of infringement (induced or otherwise) with respect to a patent on an unapproved use, as the ANDA does not induce anyone to perform the unapproved acts required to infringe.”
In another ANDA action brought prior to generic market entry, Takeda Pharmaceuticals U.S.A. v. West-Ward Pharmaceutical Corp., the patent owner argued that the generic product's proposed label would induce infringement of a patent claiming a method of treating acute gout flares by instructing that “[i]f you have a gout flare while taking [the drug], tell your healthcare provider.” 17 The patent owner argued that this instruction would “inevitably” lead physicians to use the drug for the treatment of acute gout flares, but the court concluded that it did not induce infringement, explaining that “vague label language cannot be combined with speculation about how physicians may act to find inducement,” and held that to induce infringement of a patented method, a “label must encourage, recommend, or promote infringement.” 18
In Sanofi v. Watson Labs. Inc., the Federal Circuit emphasized the requirement of purposeful action attributable to the party accused of inducing infringement, stressing that in order to induce infringement, a drug label must encourage, recommend, or promote infringement; merely providing instructions for how to use a product in an infringing manner is not enough. 19 The court noted that the term “induce” means “to lead on; to influence; to prevail on; to move by persuasion or influence.” The court also pointed out that § 271(b) explicitly includes the adverb “actively,” suggesting that the inducement must involve the taking of affirmative steps to bring about the desired result. 20
As a practical matter, the inducement to infringe provided by the proposed labeling of a generic drug in Hatch-Waxman litigation is as “artificial” as the act of infringement itself, both before and after a generic drug enters the market. This issue was addressed in GlaxoSmithKline, where the evidence clearly indicated that doctors rarely, if ever, are influenced by the labeling of the generic drug product when writing a prescription for a drug. There seems to be general agreement that doctors rarely read the labelling of branded drugs, and it's even less likely that a doctor will read the label of a generic drug and compare it to its branded counterpart. There are other sources of information, and it is these sources that influence a doctor's decision to prescribe a drug.
Furthermore, for the most part doctors do not appear to distinguish between branded and generic drugs when writing prescriptions. It is possible for a doctor to write a prescription specifying use of a branded drug, and instructing the pharmacist not to substitute a generic product, but this practice seems to be relatively rare. The premise underlying Hatch-Waxman is that FDA-authorized generic drugs are the equivalent of their branded counterpart, so it is not surprising that doctors are ambivalent as to whether a generic or branded drug is used. Patients are also unlikely to care, unless they see a chance to save money with a generic product.
Third-party payers clearly have an incentive to prefer the use of a low-cost generic rather than the branded product. The cost savings can be substantial. In GlaxoSmithKline, for example, the cost of the branded drug was reportedly $1.50 per pill, while the generic counterpart sold for 3.5 cents a pill. But decisions to purchase a generic based on cost differential are not directly influenced by the generic product's labeling. And in any event, third-party payers are not using the drug to treat a medical condition, and thus cannot be direct infringers, so they are not relevant to the question of inducement to infringe.
The use of generic drugs rather than their branded counterpart is driven not by a doctor's prescribing habits, but rather by state laws that allow or require pharmacists to substitute the generic version of a drug for a branded drug that has been prescribed, unless the physician indicates that no substitution should occur. This has nothing to do with the generic product label, which is why I assert that inducement in the Hatch-Waxman context is generally constructive, or “artificial” in the sense that the courts have used that term to describe infringement under Hatch-Waxman, rather than actual inducement.
Compounding the artificiality of the inducement to infringe provided by a generic product label is the fact that the labeling of the generic product is required by law to essentially copy the labeling of the branded counterpart. The generic manufacturer is not free to choose the language that appears on its product label; the labeling is dictated by the labeling of the branded product, which is a product of the branded company, arrived at through consultation with FDA regulators. So the content of the labeling really says nothing about the intent of the generic company. It is compelled language dictated by external sources.
Although the act of inducement is generally artificial in Hatch-Waxman litigation, there is a good policy justification for the courts' sanctioning of this doctrine of constructive inducement. Without it, method of treatment patents would be largely useless, because as a practical matter they would not be enforceable. Generic companies are “freeriders,” not only with respect to the costs of developing and testing drugs, but also when it comes to advertising and promoting drugs. And that is particularly the case prior to ANDA approval, the time at which most Hatch-Waxman litigation occurs. When method of treatment patents are asserted in the context of Hatch-Waxman, there will generally not be any marketing or promotional material attributable to the generic company on which to base a finding of active inducement of infringement. The courts are left with the proposed labeling, as described in the ANDA, and have based liability solely on this labeling. If they did not do this, and were to require actual inducement, then there generally would be no finding of induced infringement, and thus no basis to keep the generic product off the market prior to expiration of the patent.
Of course, the doctrine of contributory infringement could be enlisted, and take on a more prominent role in Hatch-Waxman litigation. So long as a branded drug company could prove that there are no substantial noninfringing uses of its product, a contributory infringement cause of action could be successful. But from the perspective of the branded drug company, a cause of action for induced infringement has much to recommend itself. If the generic product label provides instructions for using the product in a manner that would infringe the method of treatment patent, that can be sufficient for a finding of induced infringement. Proving that the drug at issue has no substantial noninfringing use could be a more difficult proposition for the branded drug company. The majority decision in GlaxoSmithKline reflects a recognition of the important role method of treatment patents play in incentivizing the development of new methods of treating diseases and the identification of new indications for a drug, important aspects of pharmaceutical innovation that can provide substantial benefits to patients. Society is well served by a patent system that supports the enforceability of method of treatment patents.
There is a significant exception to the law and regulations requiring a generic drug label to substantially copy the language of the innovator product's label, including instructions for the approved use. This exception was the subject of a decision of the U.S. Supreme Court in 2012, Caraco Pharm. Labs., Ltd. v. Novo Nordisk A/S, wherein the Court explained:
[An ANDA applicant can] submit a so-called section viii statement, which asserts that the generic manufacturer will market the drug for one or more methods of use not covered by the brand's patents. See § 355(j)(2)(A)(viii). A section viii statement is typically used when the brand's patent on the drug compound has expired and the brand holds patents on only some approved methods of using the drug. If the ANDA applicant follows this route, it will propose labeling for the generic drug that “carves out” from the brand's approved label the still-patented methods of use. See 21 CFR § 314.94(a)(8)(iv). The FDA may approve such a modified label, see § 314.127(a)(7), as an exception to the usual rule that a generic drug must bear the same label as the brand-name product, see 21 U.S.C. §§ 355(j)(2)(A)(v), (j)(4)(G). FDA acceptance of the carve-out label allows the generic company to place its drug on the market (assuming the ANDA meets other requirements), but only for a subset of approved uses—i.e., those not covered by the brand's patents.
21
A label from which a patented method of use has been “carved out” pursuant to Section viii is often referred to as a “skinny label.” The carving out of an indication from a generic label does not seem to have a significant impact on how the generic drug is used, which is not surprising, given the general irrelevance of product labeling in doctors' decisions to prescribe. And regardless of whether a doctor prescribes a generic or brand-name drug, state pharmacy laws allow or require the substitution of generic drugs with AB ratings for brand-name drugs irrespective of the indications on the label. 22 In GlaxoSmithKline, for example, a doctor testified that he “didn't actively switch” his patients to the generic product when generics launched, and instead his patients were “automatically switched” from the branded product.
Facts of Glaxosmithkline
The artificiality of the inducement that occurs in the Hatch-Waxman context took center stage in GlaxoSmithKline v. Teva. The drug at issue was carvedilol, marketed by GSK under the brand name Coreg. 23 The compound itself is covered by United States Patent No. 4,503,067, (the “composition patent”), which issued in 1985 and claims both the compound and method of using it to treat hypertension; this patent expired on March 5, 2007. 24 Carvedilol was initially approved by FDA in 1995 for use in the treatment of hypertension.
In 1997 the FDA approved a second indication for carvedilol, for use in the treatment of congestive heart failure. 25 GSK also obtained a patent on a method of using the drug to treat congestive heart failure (CHF) in 1998, United States Patent No. 5,760,069 (the “method of treatment patent”). The method of treatment patent did not claim all uses of carvedilol to treat CHF, but was limited to the use of the drug in conjunction with one or more additional active ingredients to “decrease mortality caused by congestive heart failure.” 26 GSK listed the patent in the FDA's Orange Book with use code U-233, “decreasing mortality caused by congestive heart failure.”
In March 2002, Teva applied for FDA approval of its generic carvedilol. In its Abbreviated New Drug Application, Teva certified under Hatch-Waxman's “Paragraph III” that it would not launch its generic product until the composition patent expired 2007. But Teva challenged the method of treatment patent with a “Paragraph IV” certification asserting that the patent was “invalid, unenforceable, or not infringed.” Teva received FDA “tentative approval” for this ANDA in 2004, “for treatment of heart failure and hypertension,” to become effective on expiration of the composition patent in 2007. On June 9, 2004, Teva issued a press release announcing the “tentative approval … for Carvedilol Tablets” and stating that “Carvedilol Tablets are the AB-rated generic equivalent of GlaxoSmithKline's Coreg® Tablets and are indicated for treatment of heart failure and hypertension.”
Under Hatch-Waxman, GSK could have responded to Teva's Paragraph IV challenge by filing an infringement lawsuit, which would have resulted in an automatic 30-month stay of approval of Teva's ANDA. If GSK had prevailed in such a lawsuit, by obtaining a ruling that the patent was not invalid and would be infringed by Teva's proposed generic product, the company could have been granted a permanent injunction blocking Teva from entering the market with its generic carvedilol until the method of treatment patent expired.
But GSK did not file a lawsuit at that time. Instead, on November 25, 2003, GSK filed an application with the Patent and Trademark Office (PTO) to reissue the method of treatment patent, as provided for under 35 U.S.C. § 251. As a consequence of filing for reissuance of the patent, GSK was required to surrender the original patent. The reissue patent, RE40,000, was issued on January 8, 2008 (the “reissue patent”). The claims of the reissue patent were narrowed during the course of reissuance, as illustrated by the italicized text in claim 1, which identifies the limitations added by reissue:
1. A method of decreasing mortality caused by congestive heart failure in a patient in need thereof which comprises administering a therapeutically acceptable amount of carvedilol in conjunction with one or more other therapeutic agents, said agents being selected from the group consisting of an angiotensin converting enzyme inhibitor (ACE), a diuretic, and digoxin,
wherein the administering comprises administering to said patient daily maintenance dosages for a maintenance period to decrease a risk of mortality caused by congestive heart failure, and said maintenance period is greater than six months.
On expiration of the composition patent in 2007, Teva launched its generic carvedilol with a skinny label that carved out the congestive heart failure indication, only indicating for hypertension and post-MI LVD. Teva issued press releases and marketing materials stating that its carvedilol is “an AB Rated generic of Coreg® Tablets,” but never suggested that its product should be used to treat CHF.
The FDA issued its own press release, which was published a day earlier, and went further than Teva's. It named 14 generic manufacturers, including Teva, and announced that it had approved “the first generic versions of Coreg (carvedilol).” All 14 AB-rated generics were approved based on skinny labels indicated only for hypertension and post-MI LVD. The FDA's release stated that “Coreg is a widely used medication that is FDA-approved to treat high blood pressure, mild to severe chronic heart failure and left ventricular dysfunction following a heart attack.” The FDA also stated that “[t]he labeling of the generic products may differ from that of Coreg because parts of the Coreg labeling are protected by patents and/or exclusivity.”
In 2011 the FDA required Teva to amend its carvedilol label to be “identical in content to the approved [GSK Coreg®] labeling (including the package insert and any patient package insert and/or Medication Guide that may be required).” Thus, Teva was required to amend its label to include the indication for treatment of heart failure, i.e., a “full label.” This was apparently because the original method of treatment patent was no longer listed in the Orange Book, and FDA regulations only authorize carve-out of an indication covered by an Orange Book–listed patent.
In 2014, at a time when multiple generic companies had been on the market selling generic carvedilol for years, GSK sued the two largest providers of generic carvedilol, Teva and Glenmark, alleging infringement of the reissue patent. The action against Glenmark was severed and stayed. A jury found that Teva had induced infringement of the reissue patent during the period in which it sold the product, under both the skinny label and the full label. The jury assessed lost profits damages of $234.1 million, based on the sale of a total of $74 million worth of carvedilol, a majority of which was prescribed for noninfringing purposes. The jury also found that the infringement was willful.
The district court granted Teva's motion for judgment as a matter of law (JMOL), stating that the jury's verdict of induced infringement was not supported by substantial evidence because “GSK failed to prove by a preponderance of the evidence that ‘Teva's alleged inducement, as opposed to other factors, actually caused the physicians [i.e., as a class or even at least one of them] to directly infringe,’ by prescribing generic carvedilol and to do so for the treatment of mild to severe CHF.” The district court explained that “[w]ithout proof of causation, which is an essential element of GSK's action, a finding of inducement cannot stand.”
The district court found that prior to the launch of any generic version of the drug in 2007, numerous sources of information had already informed physicians of Coreg and recommended its use for the treatment of heart failure, including the American Heart Association, the American College of Cardiology, and various publications. The court stated that GSK's Coreg label and promotion of carvedilol had already informed physicians about the uses of Coreg. Cardiologists had testified that they knew of the various uses of carvedilol before the FDA required Teva to amend its label. The court stated that “even in September 2007, when generic companies (including Teva) began selling carvedilol, doctors relied on guidelines and research, as well as their own experience, in addition to GSK marketing.” The district court concluded that “[a] reasonable factfinder could only have found that these alternative, non-Teva factors were what caused the doctors to prescribe generic carvedilol for an infringing use.”
GSK appealed, and a divided panel of the Federal Circuit reversed the district court's decision, reinstating the jury's finding of induced infringement. The majority found that the district court judge had erred in law and in fact, and that the jury's finding of induced infringement was supported by substantial evidence. It also upheld the jury's award of lost profit damages.
Writing for the panel majority, Judge Newman rejected the district court's holding that causation is a necessary element for liability for inducement of patent infringement. She also found that, even during the skinny label period, Teva had taken actions to promote the use of its generic product for the treatment of heart failure, and that it intended for its product to be used by doctors and patients in an infringing manner.
Judge Prost wrote a long dissenting opinion, in which she explained why, in her view, the district court was correct in holding that causation is a required element for a finding of induced infringement. She also expressed concern that the majority's opinion was inconsistent with Congress's intent to allow for the sale of a generic drug with approved, non-patented uses, even if other approved uses of the drug are covered by valid patent, by means of the Section viii carve-out.
Interesting Aspects of Glaxosmithkline
GlaxoSmithKline is a notable development in the law of induced patent infringement, and this section of the article identifies and discusses several particularly interesting aspects of the decision.
Citation to Supreme Court precedent
The majority decision cites two Supreme Court decisions that address the issue of induced infringement, Global-Tech Appliances, Inc. v. SEB S.A. and MGM Studios Inc. v. Grokster, Ltd., but then never explicitly applies the precedent of either decision to the case at hand. 27 However, I think it is worth spending a moment considering the implications, if any, of these decisions on the question of induced patent infringement in the context of generic drugs.
The majority cites Global-Tech for the proposition “that copying of a patented product is evidence of inducing infringement.” 28 Although court does not explicitly apply this proposition to the present case, the fact that the majority includes this statement in its decision implies that it considered it relevant to the outcome. And in fact I have seen a blog commentary in which the commentator expressed the opinion that GlaxoSmithKline's reference to Global-Tech marked a significant development in the law of induced infringement, which, if followed in subsequent decisions, would render evidence that a generic company “copied” the branded drug probative as to the question of whether the company is liable for induced infringement. Of course, the essence of Hatch-Waxman and the entire ANDA process requires that the generic drug be a “copy” of the branded product, so of necessity all generic companies copy the patented drug. The fact that copying has occurred should have no bearing on the question of induced infringement.
In my opinion, the blog commentator, and apparently the GlaxoSmithKline majority, have misinterpreted Global-Tech. In Global-Tech, the Court found “copying” of a patented product to be evidence that a party accused of inducing infringement knew that the activity it was promoting would constitute infringement of a patent. Thus, evidence of copying can be relevant in a case in which an accused party denies knowledge that a patent exists, or that the conduct being promoted will result in infringement of that patent. But this issue will rarely if ever arise in the context of Hatch-Waxman. In most cases, patents on the branded product are listed in the Orange Book, and of course the generic company has knowledge of the patents. I doubt that a generic company has ever seriously argued that it was not liable for inducing infringement because it lacked knowledge of a method of treatment patent, and it certainly was not an issue in GlaxoSmithKline.
The panel majority also cited to the Supreme Court's decision in MGM Studios Inc. v. Grokster, Ltd. for its application of “the principles of induced infringement.” Although Grokster involved inducement of copyright infringement, it is perfectly appropriate to look to the decision as precedent with respect to inducement of patent infringement. In Grokster, the Supreme Court explicitly stated that it was adopting patent law's inducement rule as a “sensible one for copyright,” implying that the Court's interpretation of the standard for finding inducement of infringement clearly to both patent and copyright. 29 Grokster is thus relevant with respect to the level of intent necessary to find Teva liable for inducement of infringement.
In Grokster, the Supreme Court found two peer-to-peer file sharing services (Grokster and Streamcast) liable for inducement of infringement based on a “record … replete with evidence that from the moment Grokster and StreamCast began to distribute their free software, each one clearly voiced the objective that recipients use it to download copyrighted works, and each took active steps to encourage infringement.” 30 Although the services could be used for noninfringing purposes, i.e., the sharing of uncopyrighted or authorized files, the evidence showed that in fact the businesses were built upon a model that relied upon a massive amount of unauthorized “sharing” of copyrighted content. There was a wealth of evidence that the filesharing companies' marketing efforts specifically targeted consumers interested in engaging in copyright infringement. The companies went so far as to provide customer support and specific instructions to users in order to facilitate the copying of works that the companies knew were copyrighted. 31
In Grokster, the Court noted that the companies were liable for inducing infringement based on evidence that went beyond mere knowledge that their products might be put to an infringing use, extending to statements and actions directed to promoting infringement. Specifically, the Court held that the necessary intent for induced infringement can be “shown by clear expression or other affirmative steps taken to foster infringement.” 32
Grokster includes an extensive discussion of the Court's earlier decision in Sony Corp. of Am. v. Universal City Studios, Inc. 33 In Sony, another decision addressing indirect infringement of copyright, the Court explicitly noted the close kinship between patent and copyright law, and indeed incorporated patent law's standard for contributory infringement into copyright law. In particular, Sony created a safe harbor from liability for indirect infringement for manufacturers who provide consumers with a product that can be used for infringing purposes, so long as that product is “capable of substantial noninfringing uses.” In Sony, the product at issue was the home video tape recorder that allowed consumers to make a library of infringing copies of broadcasted copyrighted works, but which also had substantial noninfringing uses, including “timeshifting,” i.e., recording a television broadcast to watch a more convenient time later, which the Court held was fair use and thus not infringing. In Grokster, the Court distinguished Sony by holding that the Sony safe harbor does not extend to parties that intentionally induce infringement, as demonstrated by the active steps towards promoting infringement pursued by the defendants in Grokster.
Grokster recognizes “the need to keep from trenching on regular commerce or discouraging the development of technologies with lawful and unlawful potential,” emphasizing that:
mere knowledge of infringing potential or of actual infringing uses would not be enough here to subject a distributor to liability. Nor would ordinary acts incident to product distribution, such as offering customers technical support or product updates, support liability in themselves. The inducement rule, instead, premises liability on purposeful, culpable expression and conduct, and thus does nothing to compromise legitimate commerce or discourage innovation having a lawful promise.
34
Liability even under a skinny label
To my knowledge, this case is the first in which a generic company marketing its product under a skinny label has been found liable for inducing infringement of a patent on a method of treatment carved out of the label pursuant to Section viii. 35 As Judge Prost points out in her dissent, the outcome seems contrary to Congress's purpose in enacting Section viii. 36 She expresses concern that “by finding inducement based on Teva's skinny label, which was not indicated for—and did not otherwise describe—the patented method, the Majority invites a claim of inducement for almost any generic that legally enters the market with a skinny label.” 37
In most Hatch-Waxman litigation, the infringement lawsuit is brought prior to market entry by the generic product, and the information appearing on the label is generally the sole basis for a finding of induced infringement. Without an approved generic product, there would be no press releases or product catalog listings of the type which led to a finding of liability for Teva. So this unusual outcome might be limited to cases in which the infringement lawsuit is filed after the product entered the market.
In its appellate brief, GSK argued that, regardless of the carve-out, the skinny label nonetheless included statements sufficient to induce doctors to prescribe the generic drug for the patented method of treatment. But the majority never mention this contention, instead relying solely on Teva's press releases and product catalogs. And as Judge Prost points out in her dissent, not every use of carvedilol would be infringing, since the claims included additional limitations that substantially narrow the scope of the claim. For example, the claim recites use of carvedilol in conjunction another one of the drugs recited in the claim, and must be used for treatment for a period of at least six months.
On the other hand, if GSK was correct in its assertion that the skinny label would induce infringement of its patent, then that would suggest a problem with the Section viii carve-out process itself. In negotiating the language of its label with FDA, it would seem that Teva had every incentive to avoid liability for patent infringement, and no incentive to include statements that could be interpreted as promoting infringement, since all the evidence tends to support a conclusion that differences between the information provided on a generic drug label and its branded counterpart play an insignificant role in determining whether pharmacists fill prescriptions with generic or branded products. Congress provided the Section viii carve-out for the express purpose of allowing a generic company to bring a product to market for use in unpatented methods of treatment, in spite of the fact that some uses of the drug are covered by a patent. So it is important that the FDA allow the generic company to go to market with a skinny label that clearly avoids liability.
Active promotion of infringement?
The Federal Circuit has held that, in order to be liable for induced infringement, the accused inducer must take “active” steps to promote the infringing use of the product, at times citing to the Supreme Court's decision in Grokster (discussed above). 38 In the context of Hatch-Waxman, the Federal Circuit has held that in order to induce a drug label must encourage, recommend, or promote infringement; merely providing instructions for how to use a product in an infringing manner is not enough. 39
Writing in dissent, Judge Prost argued that GlaxoSmithKline's majority holding is contrary to Federal Circuit case law, pointing particularly to Warner-Lambert and Takeda (discussed above). She found that:
Like the labels in Warner-Lambert and Takeda, Teva's label is not itself a basis for infringement. Teva's skinny label did not “encourage, recommend, or promote infringement” of the [patent]. In fact, Teva's skinny label did not even suggest the patented method; it said absolutely nothing about CHF. It is legal error for the Majority to hold otherwise.
40
Whether or not Judge Prost is correct in her assertion that the majority's decision holding Teva liable is legally erroneous, it does seem that the majority has lowered the bar in terms of what constitutes an “affirmative act” of promoting infringement. The actions of Teva can be compared with those of the defendants in the two most pertinent Supreme Court decisions, Sony and Grokster. In Sony, a company selling a product with substantial infringing and noninfringing uses was not found liable for infringement, even though the company advertised and promoted its product, and users of the product (the direct infringers) clearly understood that the product was suitable for use in copyright infringement. This seems analogous to what Teva did, publicizing the availability of its product to an audience that would know that the product was suitable for use in an infringing method of treatment. In Grokster the defendants were found liable, but as discussed above, they did take active steps to assist users whom they knew were intent upon copyright infringement. Furthermore, the businesses of the defendants were based primarily on the copyright infringement of its users. This was not the case with respect to generic carvedilol: a majority of the generic drug sold was found to be used for noninfringing purposes. The majority's decision is arguably inconsistent not only with Warner-Lambert and Takeda, but also Sanofi v. Watson Labs. Inc. (discussed above).
In Takeda, the court emphasized that while “advertising an infringing use or instructing how to engage in an infringing use” can serve as evidence of the sort of “active steps taken to encourage direct infringement” required to establish liability for induced infringement,
such instructions need to evidence intent to encourage infringement. … The question is not just whether instructions describe the infringing mode, but whether the instructions teach an infringing use of the device such that we are willing to infer from those instructions an affirmative intent to infringe the patent. Merely describing an infringing mode is not the same as recommending, encouraging, or promoting an infringing use, or suggesting that an infringing use “should” be performed.
Takeda stresses that:
This requirement of inducing acts is particularly important in the Hatch-Waxman Act context because the statute was designed to enable the sale of drugs for non-patented uses even though this would result in some off-label infringing uses. See Caraco, 132 S.Ct. at 1681–82 (“Congress understood [that] a single drug may have multiple methods of use, only one or some of which a patent covers” and that the statute “contemplates that one patented use will not foreclose marketing a generic drug for other unpatented ones.”); Warner-Lambert, 316 F.3d at 1359 (the Hatch-Waxman Act was not intended “as a sword against any competitor's ANDA seeking approval to market an off-patent drug for an approved use not covered by the patent”).
Takeda found that “[g]iven the statutory scheme[,] vague label language cannot be combined with speculation about how physicians may act to find inducement. This would seem to too easily transform that which we have held is ‘legally irrelevant,’—mere knowledge of infringing uses—into induced infringement.” 41
The knowledge element
In finding that Teva had the requisite intent to induce infringement, the panel majority appeared to find it relevant that Teva knew that doctors would prescribe carvedilol for the treatment of congestive heart failure, and that because of state substitution laws pharmacists would substitute generic product for the branded drug. 42 And presumably, Teva was more than happy to profit from the sale of generic carvedilol for use by patients who would use it in an infringing manner. But should these be relevant considerations? Of course generic companies know that an AB rated product will be subject to state substitution laws, and it is a little much to ask that company executives not be okay with that. Furthermore, courts have often stressed that the mere knowledge that a product being supplied by company is likely to be used in an infringing manner is not sufficient to constitute induced infringement, so long as the product has substantial noninfringing uses. In Warner-Lambert, for example, the Federal Circuit stated that “[e]specially where a product has substantial noninfringing uses, intent to induce infringement cannot be inferred even when the defendant has actual knowledge that some users of its product may be infringing the patent.” 43 Similarly, in Takeda, the court noted that the patent owner had conceded that mere knowledge of off-label infringing uses of the generic product would not establish inducement.
A causation requirement?
A particularly interesting question raised in GlaxoSmithKline was whether a patentee must prove that a party accused of inducing patent infringement actually “caused” direct infringement. There is no clear answer to this question, and the panel of judges that decided GlaxoSmithKline split on the issue. The question would not arise in most cases involving allegations of inducement to infringe a method of treatment patent, because most of these cases are brought under the Hatch-Waxman Act prior to any generic product entering the market, and thus prior to any possibility of causation.
As a general matter, it is well-established that one of the elements of a cause of action for inducement to infringe is a showing that someone directly infringed the patent. In Limelight Networks, Inc. v. Akamai, the Supreme Court confirmed that a person may not “be liable for inducing infringement of a patent under 35 U. S. C. § 271(b) when no one has directly infringed the patent under § 271(a) or any other statutory provision.” 44 But this element is dispensed with in most cases brought under the Hatch-Waxman Act, because Hatch-Waxman litigation generally occurs prior to generic market entry, hence the “artificial act of infringement.” 45
But GSK filed its lawsuit against Teva after the generic product was already on the market, so GSK was presumably required to establish some direct infringement of the patent. This was not an issue—clearly some percent of the generic product sold by Teva was being used in an infringing manner. But Teva argued that not only was GSK required to prove direct infringement, it had to show that it was the statements made by Teva that caused doctors and patients to use the drug in an infringing manner, as opposed to statements made by non-Teva sources. In other words, Teva argued that in order to prevail, GSK had to prove that “but-for” Teva's statements in its press releases and catalogs, the direct infringers would not have infringed.
This is an interesting argument, and a troubling one for pharmaceutical innovators, if the Federal Circuit were to hold that causation is a required element for establishing liability for induced infringement (at least outside the context of Hatch-Waxman litigation). Indeed, it was of enough concern that both the Pharmaceutical Research and Manufacturers of America (PhRMA) and the Biotechnology Innovation Organization (BIO) filed amicus briefs in the case, arguing against the causation requirement, and pointing out the harm it would cause to pharmaceutical innovators. As a practical matter, it would seem to make it very difficult to enforce a method of treatment patent once a generic product enters the market.
There are statements that can be gleaned from Federal Circuit and Supreme Court decisions that do seem to support the notion of a causation requirement for induced infringement. In briefing the issue, Teva relied primarily on a 2016 decision of the Federal Circuit, Power Integrations, Inc. v. Fairchild Semiconductor Int'l, Inc., in which the court found that the following portion of a jury instruction misstated the law of induced infringement:
[I]n order to find inducement, you must find that the party accused of infringement intended others to use its products in at least some ways that would infringe the asserted claims of the patent. However, that infringement need not have been actually caused by the party's actions. All that is required is that the party took steps to encourage or assist that infringement, regardless of whether that encouragement succeeded, or was even received.
46
The Federal Circuit held that the highlighted portion of the instruction left the jury with the incorrect understanding that a party may be liable for induced infringement even where it does not successfully communicate with and induce a third-party direct infringer. The court pointed to language from the Supreme Court decision in Global-Tech Appliances, Inc. v. SEB S.A defining the term “induce,” as it is used in § 271(b), to mean “[t]o lean on; to influence; to prevail on; to move by persuasion.” 47 Global-Tech went on to observe that each of these definitions “requires successful communication between the alleged inducer and the third-party direct infringer.” The Federal Circuit also pointed to precedent purportedly holding that “to prevail under a theory of indirect infringement, plaintiff must first prove that the defendants' actions led to direct infringement of the patent-in-suit.” 48
Although statements made in Power Integrations do seem to support a requirement of causation, they are properly viewed as dicta, and it is not clear they should be interpreted as controlling precedent. Significantly, in Power Integrations the court did not actually apply a causation requirement to the facts of that case, in the sense that it did not hold that a party accused of induced infringement was not liable because of a failure to establish causation. The court only vacated a jury's verdict finding induced infringement, it did not decide the issue of infringement. 49
Power Integrations seems to be the strongest case that either Teva or Judge Prost could find in support of this purported causation requirement, strongly suggesting that the Federal Circuit has never so held. And to my knowledge, it is an issue that is never been addressed head on by the Federal Circuit.
Of course, in GlaxoSmithKline the district court did adopt a rule that causation must be shown to establish liability for inducement to infringe, and held that as a matter of law Teva had not “caused” physicians to prescribe the drug for infringing uses, because “physicians already knew how to use carvedilol for treating CHF” prior to Teva's press releases and product catalogs, and indeed prior to the issuance of the patent. 50 But on appeal the majority reversed, finding that the district court had applied an incorrect legal standard, and holding that “precedent makes clear that when the provider of an identical product knows of and markets the same product for intended direct infringing activity, the criteria of induced infringement are met.” 51 In other words, according to the majority there is no requirement that actions of the inducer are a cause-in-fact of the infringing behavior of physicians and patients.
Judge Prost found that causation is a necessary element in a showing of induced infringement, and that Teva had not been shown to satisfy that element. 52 To the contrary, she found that “the record confirmed that doctors prescribed carvedilol according to the claimed method based on the prescribing guidelines established by the American Heart Association and the American College of Cardiology, medical research studying carvedilol, and even GSK's own Coreg® label and GSK's promotional materials advertising it. 53
Lost profit damages
The jury awarded GSK $2.41 million in lost profit damages, based on a calculation that 17.1% of generic carvedilol sales during the period of infringement were used by patients for the method claimed, and that GSK was charging at least $1.50 per pill. Teva only sold $74 million worth of carvedilol during the allegedly infringing period, for all indications (most of which are unpatented), at a cost of less than 3.5 cents per pill, and reportedly lost money in the process.
On appeal, Teva argued that the jury should not have awarded lost profit damages, because under established Federal Circuit law an award of lost profit damages must be based on “but-for” causation. In other words, a patent owner should only be awarded “lost profits” damages if it can prove that “but for” the infringement, it would have earned that amount of profits. Teva argued that with at least seven other generic manufacturers selling generic carvedilol (most of whom were not sued by GSK), if pharmacists hadn't filled prescriptions with Teva's generic product, they would have filled prescriptions with some other generic version of carvedilol, not the branded product.
This certainly would seem to be the case, but the Federal Circuit rejected this argument, agreeing with the district court's explanation of the law, as set forth in the following jury instructions:
The undisputed evidence is that [Teva's] generic carvedilol is interchangeable with the generic carvedilol of the non-party manufacturers; therefore, the generic carvedilol of these non-party manufacturers is an
The majority concluded that the district court correctly recognized that “[i]t doesn't matter whether the
Undue deference to a jury decision?
GlaxoSmithKline might be explainable as simply an example of a Federal Circuit panel showing a high degree of deference to a jury decision. For example, at one point the majority remarked that:
[O]ur colleague in dissent applies an incorrect standard of review, for this court on appeal of a jury verdict does not find facts afresh, contrary to the substantial evidence standard. For example, the dissent finds that neither “Teva's press releases [nor] its product catalogs encourage doctors to practice the patented method.”
56
Thus, it seems that the majority based its decision on an assumption that Teva's press releases and/or its product catalogs did, as a matter of fact, encourage doctors to practice the patented method. If we are going to accept as a given that Teva's materials did encourage doctors to practice the patented method, then liability for inducement to infringe surely follows. However, I think many would agree with the dissent, in her implicit conclusion that no reasonable juror could find that either the press releases or product catalogs could have encouraged doctors to practice the patented methods, under the correct legal standard. The majority states that the jury's factual conclusion that Teva's materials encourage doctors to practice the patented method is supported by testimony that doctors “do read press releases and product catalogs.” But even if we are to assume that doctors read the press releases and product catalogs of generic companies, reading these materials doesn't necessarily mean a doctor will be “encouraged” by them to prescribe the generic drug for a patented method of treatment.
Final Thoughts on Glaxosmithkline
The outcome in GlaxoSmithKline appears to be inconsistent with the intent behind Hatch-Waxman in at least two respects. First, it results in a finding of liability for patent infringement based on the sale of a generic product under a skinny label. Congress clearly intended the Section viii carve-out to enable generic products with substantial noninfringing use to enter the market, regardless of the fact that other uses of the product are still protected by patents, and would not have anticipated that a generic company would take advantage of the process and still be held liable based on public statements announcing the availability of an AB-rated product. And second, it results in a finding of liability for infringement of a patent by a generic company that, in the words of Judge Prost, “did everything right,” waiting out the composition patent and successfully challenging the method of treatment patents that were listed in the Orange Book at the time the company filed its ANDA and launched its generic product.
In her dissent, Judge Prost voiced concern that the decision would discourage generics from entering the market, finding that:
The record shows that Teva did everything right—proceeding precisely as Congress contemplated. Teva launched its low-cost generic carvedilol for unpatented uses using a skinny label. And Teva did not encourage doctors to use generic carvedilol to practice the one still-patented use.
57
In briefing the case, Teva argued that “what GSK really wants is to impose an ‘unprecedented’ and unfounded duty on generic drug companies—to go beyond ‘carving out’ patented indications by affirmatively ‘disclaim[ing]’ them.” GSK reportedly asked the district court to make clear that Teva was obligated to provide “disclaimers clarifying its product was not approved for heart failure.” 58
However, it is far from clear that the decision will substantially discourage generics from entering the market. Even GSK seems to acknowledge that a generic company will not be liable for induced infringement if it affirmatively disclaims infringing use of its generic product. And all the evidence I have seen suggests that an affirmative disclaimer of this sort would have a negligible impact on sales of the generic product, because it would generally not influence the prescribing decisions of doctors, nor would it affect generic substitution by pharmacists, a practice allowed for or even required by state law.
But if it is in fact the case that generic companies can bring a generic product like carvedilol to market, i.e., a product with both substantial infringing uses as well as substantial noninfringing uses, so long as the company makes effective use of Hatch-Waxman's Section viii carve-out provision and includes a disclaimer in its press releases and product catalogs, then how valuable are method of treatment patents? If method of treatment patents are effectively unenforceable with respect to a generic drug that has substantial noninfringing uses, then how do we incentivize branded drug companies to seek out and develop new indications for their drugs?
It seems to be a Catch-22 situation. If we allow branded drug companies to block generic market entry even after patents on the drug itself have expired, and the drug has substantial non-patented uses, then this thwarts the congressional intent behind the Section viii carve-out. But if we allow the generic drug to enter the market in spite of the fact that some of it will be used in a manner that infringes the method of treatment patent, then this reduces the incentive for the development of new indications. As long as we have state substitution laws that actively promote the infringement, I do not see how the Catch-22 can be resolved. In reality, it is not the generic companies that are inducing infringement—they do not need to. It is the current system of drug distribution, and particularly state substitution laws. But this is not the sort of inducement amenable to resolution by patent infringement litigation.
A discussion of possible solutions to the Catch-22 would go well beyond the scope of this article. At a high level of abstraction, I can imagine a reformed drug distribution system in which pharmacists would only provide a generic version of a drug to patients when that drug has been prescribed for a noninfringing use. In cases where the prescription is for a patented method of treatment, the pharmacists would only be allowed to fill the prescription with branded product, thus allowing the innovator company to profit from sales directly attributable to the patented method of use. But given the current emphasis on bringing down the cost of drugs, that sort of reform seems unlikely in the foreseeable future. Another approach would be to consider supplementing the currently available three-year extension of marketing exclusivity for drugs that have been the subject of a new clinical investigation. 59