Riddles and Lessons from the Prescription Drug Wars: Antitrust Implications of Certain Types of Agreements Involving Intellectual Property


Remarks presented before the Attendee of The ABA "Antitrust and Intellectual Property: The Crossroads" Program

San Francisco, California

Sheila F. Anthony, Former Commissioner

Good afternoon, I'm very pleased to be here. I'd like to thank Howard and the ABA for inviting me. It's an added bonus that the ABA chose to hold this conference in beautiful San Francisco. I love this city. The FTC has a regional office here, so I have a great excuse to visit often.

I began the practice of law as an intellectual property lawyer, principally in trademark and copyright. I am proud to say that, among other things, I helped to protect the good name of Wisconsin University's mascot, Bucky the Badger, from being sullied by those who sought to use his likeness to sell condoms. Their advertising catch phrase was "Get Lucky with Bucky!" Bucky's plight, along with the many other cases I handled during that phase of my career, left me with a deep appreciation of the important role that the protection of intellectual property rights plays in the functioning of strong markets. The need to differentiate products, to control the marketing of products, and to keep others from unfairly infringing on products are sine qua nons of a healthy marketplace. Perhaps most important of all, is the role that intellectual property rights play in spurring innovation.

As an FTC Commissioner, I am now forced to add a new set of considerations and goals to my thinking about intellectual property issues. Of course I'm speaking of those embodied in the antitrust laws. Antitrust, like intellectual property, prizes innovation. However, the two disciplines can sometime fall out of harmony when issues of exclusion arise. Striking a balance between exclusionary rights necessary to fuel innovation and those that unduly dampen competition is an ongoing challenge, and one that lawyers from both legal realms must take on together. While it has been said many times, it bears repeating, especially in the midst of the new technology economy that is raging all around us: antitrust and intellectual property lawyers need to take off the blinders imposed by their own disciplines and broaden their perspectives to include each other's concerns and goals. Such a broader perspective is crucial to effectively serve clients, and in the larger scheme of things, to create effective rules governing technology markets going forward.

As always, I must emphasize that the views I express here today are my own, and do not reflect those of the Commission or any other Commissioner. While written copies of my remarks are not available for distribution today, I plan to post this presentation on the Commission's web site, ftc.gov, within the next week or so. I also commend you to the web site if you are looking for official documents relating to the cases I discuss today. I see that Howard has kindly placed print outs of some of those documents in the materials.

Today I would like to talk about two matters that have come before the Commission within the last few months that challenged me to sort through the thicket that often is found at the intersection of the intellectual property and antitrust laws. The first, called Abbott/Geneva, involves an agreement between a branded drug manufacturer and a generic manufacturer whereby, as our complaint charged, the branded company paid the generic to stay out of the market. The second, Lilly/Sepracor, involved a licensing agreement whereby the manufacturer of a leading branded antidepressant medication sought to acquire the intellectual property rights to a successor product.

I will describe the facts of these cases, my concerns about the conduct at issue, and their ultimate dispositions. I will also identify some of the factors that would likely trigger my concern as an enforcer in future cases presenting similar scenarios. I should warn you at the outset that along with providing some insights into the outcomes in two instances where the antitrust rubber met the intellectual property road, these matters suggest many questions that merit further thought. I've been wrestling with these broader questions quite a bit, and now, I invite you to think about them too.

Let's start with Abbott/Geneva.

This case was considered by the Commission along with a companion case involving Hoechst Marion Roussel and a firm called Andrx. The two cases were considered together due to some overarching similarities in the issues presented and because their investigatory phases came to a conclusion at around the same time. Abbott/Geneva was settled by consent order prior to the commencement of litigation. The Hoechst/Andrx case, however, did not settle. Rather, the Commission voted out a complaint, and the case is now in administrative litigation. I am not at liberty to discuss the Hoechst/Andrx matter in any detail because, following the administrative trial, the matter can be appealed to the Commission for review. I will instead focus on the Abbott/Geneva case, and would like to make it clear that, despite some superficial similarities, such as that both involve the interplay of the Hatch-Waxman generic drug act and private agreements between branded companies and generic manufacturers, the two cases are factually distinct. In other words, each stands or falls on its own merits, and the legal conclusions applicable to one may not hold for the other.

Before describing the terms of the agreement at issue in Abbott/Geneva, a few words about the Hatch-Waxman Act are necessary to provide context. One of the aims of the Hatch-Waxman Act is to get generic drugs to market earlier. It does this by both streamlining the FDA's generic approval process and encouraging generic manufacturers to challenge weak patents. Under the Act, the FDA is empowered to approve a generic for market even before the branded company's patents expire, if, along with filing an "Abbreviated New Drug Application," the generic manufacturer certifies that any otherwise blocking patents are invalid or not infringed. The Act provides a powerful incentive to challenge patents by granting a 180-day market exclusivity period to the first filer. Under this provision, the FDA cannot approve any other generic to enter the market until the first filer has had its product on the market for 180 days. When a generic filer contends that the branded company's patent is invalid or not infringed, it is required to give notice of this contention to the branded company. If, within 45 days of such notice, the branded company files suit, the Act provides for an automatic stay during which time the FDA may not grant final approval. The stay is effective until the shorter of: the patent expiration date, a final court ruling in favor of the generic, or 30 months.

Turning now to the specifics of the Abbott/Geneva matter, as alleged in the Commission's complaint, the prescription drug in question is Abbott's Hytrin, the brand -name for terazosin HCL. It is used to treat hypertension and enlarged prostate. Both of these chronic conditions affect millions of Americans each year, many of them senior citizens.

In January 1993, Geneva filed for FDA approval of generic terazosin HCL in tablet form; Geneva filed a similar application for a generic version of terazosin in capsule form, in December 1995. In April, 1996, Geneva filed a certification with the FDA representing that its products were not blocked by the Abbott patent because that patent was invalid.

Within the 45-day period, on June 4, 1996, Abbott sued Geneva, claiming patent infringement by Geneva's generic terazosin HCL tablet product. The suit triggered the 30-month stay of final FDA approval of the tablet, until December 1998. However, Abbott inadvertently failed to file a similar lawsuit regarding Geneva's generic capsule version of the product, even though both tablets and capsules involved the same potential infringement issues. Therefore, the FDA's review and approval process regarding the capsule continued.

The FTC's complaint alleged that Geneva, confident that it would win its patent infringement dispute with Abbott, planned to bring its generic terazosin HCL capsule to market as soon as possible after FDA approval. As the first filer on capsules, Geneva would enjoy the 180-day exclusivity period.

In late March, 1998, When Geneva finally received FDA approval to market its unchallenged generic capsules, Geneva informed Abbott that it would launch its product unless Abbott paid it not to enter. Abbott estimated that the entry of a generic would cost $185 million in sales in the first six months. And, guess what happened? Abbott and Geneva made a deal.

Geneva agreed not to bring its generic product (in either capsule or tablet form) to market until the earlier of: 1) final resolution of the suit on the tablets, including appeal to the Supreme Court level, if necessary; or 2) entry into the market of another competing generic product. Geneva also agreed not to transfer, assign or relinquish its 180 -day exclusivity right to market its generic product.

In exchange, Abbott agreed to make nonrefundable payments to Geneva of $4.5 million per month until the district court ruled. If the ruling was in Geneva's favor, Abbott agreed to put $4.5 million monthly into an escrow account during the appeal process. The winner on appeal would receive the escrow funds. The trial court was not notified of the agreement.

Living up to its agreement, Geneva did not introduce its generic capsules after FDA approval in late March, 1998, and in April, 1998, began collecting its $4.5 million per month, an amount, by the way, that exceeded what Abbott estimated Geneva would receive from actually marketing the drug. The following September, the district court ruled in Geneva's favor, invalidating Abbott's patent. Despite this victory, and the subsequent expiration of the Hatch-Waxman 30-month stay in December, 1998, Geneva still did not enter the market with either its generic tablet or capsule, content to have Abbott make monthly $4.5 million payments into the escrow account. On July 1, 1999, the Court of Appeals for the Federal Circuit affirmed the lower court's decision. Under the agreement, Geneva was to await Supreme Court consideration before entering; however, with the FTC's investigation pending, it canceled the agreement and finally entered in August, 1999.

That's the conduct in a nutshell. At base, we have an agreement between competitors whereby one is paying the other to stay out of the market. Geneva's agreement to neither use nor relinquish its 180-day exclusivity rights also served to block any other potential entrants.

Some might argue that this agreement amounted to nothing more than a stipulated preliminary injunction, with the equivalent of a bond for potential damages. However, upon examination, this agreement goes far beyond the terms of a PI. Allow me to highlight some of the major differences.

First, a PI only lasts through the litigation at the trial court level, but this agreement kept Geneva from entering the market until appeals through the Supreme Court level had been completed.

Second, a PI may require the moving party to post a bond to cover any damages should the enjoined party prevail, but this agreement had a monthly, non-refundable $4.5 million payment through the trial court level. At the appeals levels, Abbott made payments to an escrow account, but their effect on Geneva's incentives to delay entry was similar.

Third, a PI would have restrained sales of the allegedly infringing product, but this agreement restrained Geneva from selling anyterazosin in the United States, even if it developed a non-infringing product.

Fourth, the agreement restrained Geneva from transferring or relinquishing its 180-day exclusivity right. This kept any other would-be entrants from coming to market.

Finally, and perhaps most importantly, there was no judicial review of this agreement. Indeed, there was no case at all formally pending on the capsule form of the drug. A judge's review, which among other things, takes the public interest and likelihood of success on the merits into account, distinguishes this private agreement from a publicly scrutinized preliminary injunction.

In a situation like that presented in Abbott/Geneva, the monopolist and potential entrant face a great temptation to maximize financial rewards by delaying generic entry. The monopolist is happy to share a portion of its monopoly rents in order to maintain its monopoly. And the would-be entrant is happy to receive as much, or potentially even more money, than it would by entering.

Moreover, with purely private agreements, that is, those lacking any court review, the temptation is to reach an agreement at the public's expense. The patent system, however, exists for a reason. Patent litigation, which tests and reveals the validity of a patent, should be allowed to run its course. A preliminary injunction issued under Rule 65 of the Federal Rules of Civil Procedure is the proper vehicle for maintaining the status quo while the complicated infringement issues are resolved by the court. An actual PI, of course, would likely be much more limited in its terms than the Abbott/Geneva agreement. For example, while the court typically requires the posting of a bond, in most instances, payment would be limited to actual damages and would not be forthcoming until the case had run its course.

Allow me to change to my defense lawyer hat for moment and mention some potential arguments in favor of private settlement agreements of this sort. First, settlements are generally favored in our judicial system. They save judicial resources and often yield quick resolutions. Second, some restrictions beyond the general prohibition against marketing the allegedly infringing product may, in certain circumstances, be appropriate to make the injunction fully effective. Third, if the alleged infringer is potentially judgment proof or failing, wouldn't up-front payments be necessary to keep the infringer from making a desperate entry attempt for which the patent holder cannot be compensated at the end of the day?

Taking these in reverse order, let me respond to each. The case of the potentially judgment-proof, would-be entrant seems to be best addressed by court intervention. Courts considering PI motions weigh the likely effects of the injunction on the party enjoined, the likelihood of success on the merits, the public interest, and the irreparability of the harm. If the would-be entrant lacks the resources to pay the potential damages, and the moving party is likely to succeed on the merits, it seems reasonable to believe that the injunction will issue.

Next, ancillary restraints to prevent circumvention of the injunction might be appropriate in some cases, but again, judicial review would help a lot in this regard. A judge could distinguish necessary and appropriate ancillary terms from those that are overreaching and contrary to the public interest, including the public's interest in vigorous competition.

Finally, while settlements are generally favored, we cannot overlook other interests and concerns. Settlements can reduce costs and, through licensing or similar means, even speed and engender competition. On the other hand, settlements between monopolists and would-be entrants are ripe for collusive dealings that leave consumers and competition behind. In short, the public's interest must be represented at the settlement table. This is best left to a court. Companies might also seek guidance from the antitrust enforcement authorities, like the FTC.

The FTC's complaint alleges that the agreement between Abbott and Geneva constituted an unreasonable restraint of trade; that Abbott monopolized the relevant market; that Abbott and Geneva conspired to monopolize the relevant market; and that the acts and practices are anticompetitive in nature and tendency and constitute unfair methods of competition, all in violation of Section 5.

Under the terms of the settlement, Abbott and Geneva are barred from agreements where first-filers agree not to give up or transfer their 180-day exclusivity rights or not to bring non -infringing drugs to market. In addition, agreements where the generic is paid to stay off the market would have to be approved by the court when undertaken during the pendency of patent litigation, and the companies would be required to give the Commission 30 days' notice before entering into such agreements in other contexts. In addition, Geneva waived its exclusivity relating to tablets; this allowed immediate entry by other generic manufacturers of the tablet form of the drug.

Generalizing for a moment, allow me to point out some of the factors that ought to raise red flags in your mind if you encounter them in your daily practices. First, beware of preliminary injunction type agreements which include provisions that go beyond Rule 65. For example, up-front non-refundable payments from a patent holder to an alleged infringer. Second, proceed with caution when considering private preliminary or permanent settlements of patent litigation that are not reviewed by a Court. Without review, weak patents may be honored in favor of a share of the monopoly rents at the expense of the public interest. Third, and more particularly, do not succumb to the temptation to overreach by including unnecessary terms like those in the Abbott/Geneva agreement. There, Geneva agreed not to market non-infringing products or relinquish its right to the180-exclusivity period.

As a final note on this case, I would like to reiterate a point included in the unanimous Commission statement accompanying the Abbott/Geneva consent orders. As cases of first impression, future conduct prohibitions alone were appropriate, but "in the future, the Commission will consider its entire range of remedies in connection with enforcement actions against such arrangements, including possibly seeking disgorgement of illegally obtained profits."

I would now like to turn to the Lilly/Sepracor matter.

Throughout 1999, the Commission conducted an investigation of Lilly's announced intent to acquire an exclusive license to Sepracor's intellectual property rights in a new drug, called R-fluoxetine, a close relative of Lilly's popular antidepressant drug, Prozac. Lilly's Prozac patents are set to expire in 2004, while Sepracor's R-fluoxetine patent will not expire until 2015. This scenario raised various concerns regarding horizontal competition between Prozac and R-fluoxetine and the potential impact on soon-to-enter manufacturers of generic Prozac. After a thorough investigation and a lot of thought about complicated I/P and antitrust issues, in April, the Commission voted to close the investigation and allow the transaction to proceed unchallenged.

Now, why, you may ask, am I discussing a matter that did not result in a complaint being filed? I firmly believe that the Commission needs to talk about not only the actions it takes, but about those it does not. Such discussions are particularly instructive where similar future scenarios, differing only in the factual details, might lead me to authorize a challenge.

At base, Lilly/Sepracor examined the competitive implications of a market leader buying the rights to a potential successor product. Some complaining parties asserted that Lilly's purchase was an attempt to, in-effect, unfairly extend the Prozac patent by 11 years - sort of like moving the goal line just as your opponent is about to score.

Before delving into the intellectual property and competition implications of deals like this, let me back up for just a moment to provide some background. Depression is a heterogeneous illness that manifests itself in a variety of ways. While it can be a one time event, it is often a recurring chronic condition. Treatment can be further complicated by the varying reactions that different patients have to the available antidepressant drugs.

Although other, older types of drugs are still used in some cases, today's most commonly used treatments are a set of drugs called "selective seratonin reuptake inhibitors" or SSRIs. The main SSRIs include Prozac, Zoloft, Paxil, and Celexa. Even though they are all SSRIs, they have different clinical profiles in terms of effectiveness for a given patient, side effects, etc. By most measures, Prozac leads the market for antidepressants.

The chemical name for Prozac is fluoxetine. Like many drugs, fluoxetine is a mix of two distinct isomer molecules. The isomers contain the same atoms, but those atoms are arranged spatially as mirror images. One is called S-fluoxetine and the other R-fluoxetine. Scientists have found that, in some cases, one isomer has the medicinal effect, while the other is primarily responsible for the side-effects. Thus the interest in manufacturing and testing the individual isomers.

Sepracor took up this research challenge and now holds the patent on R-fluoxetine. Sepracor is a relatively small drug company. Its business model is based primarily on research and licensing, rather than on marketing drugs itself. Testing is not yet complete for R-fluoxetine, but the drug might be similar to Prozac, and yet avoid some of Prozac's side-effects such as sexual dysfunction and insomnia.

So, what are some potential concerns that an acquisition like this raised in my mind?

First, I had the general concern that arises when a dominant firm buys a potential competing product. The easiest way to appreciate this concern is to ask yourself whether competition between the dominant firm's product and the new product would be stronger if a separate and unrelated company were marketing the new product. As a general matter, all else being equal, it seems like competitive vigor is likely to be stronger if the products are in separate hands.

Second, I was concerned about the competitive implications of a potential share-shifting strategy. For example, with the rights to R-fluoxetine, Lilly could introduce R-fluoxetine in the waning years of the Prozac patent and attempt to market it in such a way as to move its Prozac share to the new drug. Since the new drug is patent protected, this might shield Lilly's overall market share from generic competition.

You might ask: in light of these concerns, why didn't I vote to authorize a complaint? While I cannot share the details of confidential factual information gathered in the course of this non-public investigation, I can tell you about my thinking on some of the broader legal issues raised by the case.

Let me start with the share-switching concern. The theory is founded on the premise that R-fluoxetine will prove to be essentially the same as Prozac - that it will offer few new benefits - and that patients and their doctors will nevertheless be persuaded to use the new drug rather than using the presumably cheaper generic form of Prozac. Now, in 2004, upon expiration of the Prozac patent, generics will come on the market, and prescriptions for Prozac can be filled by substituting the cheaper generic, but prescriptions for R-fluoxetine will have to be filled with the more expensive, name-brand drug. While there may be circumstances where a scenario like this presents a real risk, as a general matter, I'm inclined to trust doctors and patients to determine the relative worth of a new product.

This case could also be analyzed under a more generalized horizontal competition theory. As many of you know, market definition is often a critical factor in horizontal antitrust cases. If the market is defined narrowly, and concentration is high, and if the acquiring firm is dominant and likely to remain so, the case against a transaction is likely to be stronger than if the market is defined more broadly and concentration is lower.

In the Lilly/Sepracor matter, the market picture was ambiguous. In a narrow market limited to Fluoxetine and its derivatives -- that is, Prozac, generic fluoxetine, and R-fluoxetine - concentration would be high. Under this market definition, Lilly, the dominant player, would be proposing to buy one of its closest competitors, and in doing so would prevent a major new competitive force from coming in.

However, the market could also be defined more broadly. If R-fluoxetine is sufficiently different from generic and branded Fluoxetine, the market might include all SSRIs. And with Zoloft, Paxil, Celexa, and other SSRIs included in the market, concentration goes down and Fluoxetine's future market dominance is less certain. In this broader market, Lilly is competing with many more players and is no longer as plainly dominant.

I should emphasize that antitrust enforcers often must act without perfect information. But just because the potential market parameters may be uncertain prior to litigation, this by no means translates into a free pass for a proposed deal. That's what expert witnesses are for, after all. As a general matter, however, as concentration and dominance decrease, so too do my concerns about horizontal anticompetitive effects.

Another consideration in cases like Lilly/Sepracor, as in many acquisitions, is efficiencies. If the acquiring firm, for example, can get the new product to the market faster than other companies, or market it better than anyone else, these factors might weigh in favor of the acquisition in my mind. An intellectual property acquisition might also be efficient if it defuses a legitimate blocking patent suit and thereby avoids the delays inherent in such litigation. These speed-to-market type efficiencies are potentially heightened in the pharmaceutical context where new drugs sometimes assume life-and-death importance.

I don't want anyone thinking that efficiencies are a magic bullet, however. While it may be axiomatic, I'd like to make it clear that while I weigh cognizable efficiencies against potential anticompetitive effects, efficiencies, standing alone, do not somehow immunize an otherwise anticompetitive transaction from challenge.

The big pictures issues are: where should we draw the antitrust lines to preserve incentives to innovate, and, conversely, where should we draw the intellectual property lines to not unduly affect competition? These questions are difficult to ponder in a vacuum. A real world deal like Lilly/Sepracor, and the real world facts that come along with it, give us some clues as to the questions we need to ask, as well as some hints about what the right answers may be.

For example, should there be a blanket prohibition against dominant firms acquiring exclusive rights to follow-on technologies? What effect would such a rule have on innovation?

It seems to me that a blanket prohibition could advance innovation in some cases and impede it in others. If the dominant firm knew that it could not buy innovations, it would have an incentive to develop them, itself. On the other hand, companies like Sepracor might not risk capital to develop follow-on technology if they could not sell it to the originator, who is often the party most interested in the follow-on. Given these potential contradictions, its seems that the most prudent course is to examine matters on a case-by-case basis.

On the intellectual property side, Lilly/Sepracor makes me ask whether aspects of the patent system might allow perpetual exclusivity. Should we consider changes that might limit the potential to "game" the system?

Clearly, finding the right balance between competitive concerns and innovation concerns is an ongoing challenge.

Let me conclude by saying that I realize I may have raised more questions than I've answered. In my experience that's unavoidable when crossing through the intersection of antitrust and intellectual property law. I'm counting on you learned folks out there to think hard about these issues and to share your perspectives and insights with the rest of us. I said it at the outset and I'll say it again: lawyers and economists in both the antitrust and intellectual property fields need to keep each others' aims and concerns in mind as we go about our day-to-day business so that our common aim of competitive and innovative markets can be realized.

Thank you very much.