HLS Seminar discussion on “Pay-for-delay settlements”
We start the series of posts from Harvard Law students with a great introduction to the antitrust issues that arise in connection with “pay-for-delay settlements”/”reverse payments” in the pharma industry. The post is authored by Paul B, and builds on the relevant readings in the syllabus. We have also included other students’ reactions in the “comments section”.
For those short on time and already familiar with the topic, go ahead and click on the link below to skip to “Questions for discussion”.
Pay-for-delay (PFD, “reverse payments”)
The issue is tricky because it lies at the intersection of patent, food/drug and antitrust (AT) law, and its unclear to which of these we should look to address abuses that arise from the US generic drug regime. In short, when a pharmaceutical company develops a new (branded) drug, it first seeks a patent. The initial problem is that the PTO grants patents fairly generously, in a largely non-adversarial process, so in many cases the branded drug will be patented even though it is arguably not novel, non-obvious, etc. The drug then goes through a lengthy and expensive testing and Federal Drug Agency approval process (a New Drug Application, or NDA), which may eat up a sizable share of the patent protection period.
Once the drug is FDA approved and hits the market, the Hatch-Waxman amendments to the Food, Drug and Cosmetics Act kick in. Consistent with the themes we’ve discussed throughout the term, Hatch-Waxman attempts to balance the fostering of innovation (by protecting the patent-granted monopoly for truly innovative new drugs) against the desire to foster competition by allowing low-cost generics on the market as soon as possible. For a normal (i.e., non-pharma) patent, the way to challenge a disputed patent would be for an alleged infringer to place his product on the market, and for the patent holder to sue for infringement damages and an injunction against future sales. If the parties settle, the infringer might pay the patent holder part of the alleged damages (a higher share the more likely they are to get an adverse verdict, based on the probability that a court will find that the disputed patent was both valid and infringed by the defendant) and/or there may be some sort of licensing or contract manufacturing agreement. Such agreements typically do not raise serious AT concerns.
In the case of pharmaceuticals, by contrast, Congress decided in Hatch-Waxman for various reasons to set up a regime in which the legal challenge comes before the infringement. So a company which develops a generic version of a branded (and patented) drug begins by filing an abbreviated new drug application (ANDA), which is much easier to approve than an NDA (the company must only show that the drug is bioequivalent to the branded drug). As part of the ANDA, the generic company informs the branded drug manufacturer that it intends to challenge the legitimacy of its patent. Assuming the branded company wishes to defend its patent and challenge the ANDA, a 30-month delay is automatically imposed before the generic can go to market, during which the companies may litigate the claim. If (as happens surprisingly often) the generic wins, it is granted a 6-month exclusivity period to market its generic version (creating a market duopoly) before other generics may enter the market. During that period, the generic will typically price its drug below the price point of the branded drug (which has been charging the monopoly price) but well above the competitive market price which will obtain once other generics enter the market (roughly 15% of the monopoly price, on average). This system (1) rewards the first firm to challenge potentially weak patents which are wrongly imposing monopoly pricing on consumers (2) allows the issue to be resolved prior to costly commercialization of a potentially infringing product, (3) preserves and expedites the patent monopoly of truly innovative drugs, and (4) ensures that market pricing is achieved within 4 years of the filing of a legitimate pharma patent challenge.
UNLESS, the parties settle. Here, because no infringement has yet occurred, proper settlement damages will in theory be “reverse”; that is, if there is a 50% likelihood that the generic has been kept off the market by an invalid patent, the branded drug holder may offer to pay the generic 50% of what it could have made by marketing the drug during that period (rather than the normal process of the infringer paying the patent-holder 50% or some other share of what it actually did earn from infringing). The concern here is that both parties have an incentive for this payment to reflect more than just their best estimates of patent validity, damages or litigation costs (all legitimate considerations in a settlement), but rather to split up the monopoly profits. That is, if there are 6 years remaining on the branded drug’s patent, and the parties agree it is 50% likely that the patent is invalid, they could agree to a settlement that the generic would just wait 3 years to enter the market. When the branded company instead pays the generic “reverse damages” in return for an agreement to stay off the market for the full 6 years, there is a concern that the firms are essentially maintaining a bogus monopoly at the expense of consumers. If, say, BrandX sells for $100 per pill, and the market price under full competition is $20, brand and generic may agree to a pay-for-delay settlement in which brand pays generic $30 for each unit of BrandX sold for the remainder of the patent life. This allows them to split monopoly rents: brand makes $70 per pill, still well above the market price, for a drug that arguably should not have patent protection, and generic earns $30 per pill for doing nothing, much better than it could have done at market. This same sort of agreement can be done with generic performing some contract manufacturing for brand, also at those prices.
Questions for discussion
The basic question is whether these reverse payment agreements, where payments from branded to generic may reflect an effort to maintain and share an improperly obtained monopoly, rather than just shared expectations regarding the costs and expected outcome of litigation, should be barred under the AT laws.
1. Are reverse payments even undesirable? There is an argument that they are not. If patents are presumed valid, then patent holders should have the choice to litigate to protect their patents or settle to avoid litigation hassles, as long as the monopoly is not extended beyond the scope of the patent. The judicial system as a whole benefits from settlements, and the entire purpose of settlement is defeated if a third party can assert (treble damage) claims and essentially demand a mini-trial on the merits each time there is a settlement.
2. If some reverse payments are anticompetitive, which ones? There are many options here, falling on a spectrum:
*Federal, Second and Eleventh Circuits: PFD agreements can only be illegal under unusual circumstances, such as where the agreement exceeds the scope of the patent’s protection; the settlement was part of a fraud; or the underlying patent litigation was objectively baseless.
*Current DOJ view: a plaintiff may establish a prima facie case of liability by showing a payment to the generic accompanied by the generic’s agreement to withdraw its patent challenge. Defendants may then rebut the presumption by showing reasonableness, such as a demonstration that the payment amount did not exceed litigation costs. Alternatively, defendants can show that the ‘settlement preserved a degree of competition reasonably consistent with what had been expected if the infringement litigation went to judgment.
*S. 369 treats PFD settlements as presumptively unlawful, except when the settlement parties can show pro-competitive benefits outweighing anti-competitive effects by clear and convincing evidence. To overcome the presumption of illegality, parties of a proposed agreement may present evidence to support procompetitive effects of the agreement. In examining whether the parties have met the burden of proof, the fact finder may consider seven factors listed in the bill.
*HR 1706/3962 proscribes settlements where a generic supplier would receive ‘anything of value’ from the branded drug manufacturer (beyond the right to enter the market before the exclusivity period of the patentee expires) and where the generic drug manufacturer also agrees to stop researching, developing, manufacturing, or selling the generic drug.
So some relevant questions to consider here are: (1) is there ever a good (i.e., procompetitive) reason to have a reverse cash payment plus a longer delayed entry period, rather than just agreeing to let the generic enter the market sooner without any reverse payment? (2) how are we to assess terms other than paying and delaying, which often accompany PFD settlements, such as contract manufacturing, licensing, etc? Should we assume that these are merely clever pretexts for cash payments to generics, or might there be procompetitive terms we want to allow? (3) If PFD is not to be per se illegal, which factors are most important in assessing AT challenges? Should we assume that any dollar value above a certain level (e.g., expected litigation costs) is presumptively illegal? Is the prima facie strength of the patent the key factor? How important is it whether there are other generics waiting to enter the market?
3. Relatedly, how much discretion should government bodies (FTC, DOJ, FDA) have in deciding which settlements to challenge/approve? Does the recent change in DOJ’s position under Obama just go to prosecutorial discretion, or is there an admin claim here that the issue of drug pricing is so heavily politicized that we should keep decisions in individual cases away from the elected branches and in the courts? Should there be a private right of action in challenges to PFD agreements (proposed legislation varies in this regard)?
4. Under what standard (per se, rule of reason, quick look) should challenged PFDs be assessed? Does the novelty and technical nature of these agreements argue for more detailed case-by-case evaluation, or does that defeat the purpose of settlement?
5. Which courts should hear the challenges? If the ultimate question is whether a settlement reflects the actual strength or weakness of the underlying patent claims, does that mean that the Federal Circuit (which hears all substantive patent claims) should decide all these cases?
6. Should the issue be resolved by Congress rather than the Supreme Court, and if so, should Congress tweak the innovation/competition balance by revising Hatch-Waxman and/or imposing a more adversarial patent approval process for drugs, rather than by imposing per se AT liability? In other words, are problems inevitable if we wait to address a potentially invalid patent until the point in the process at which a drug has already been broadly commercialized at monopoly prices? Are antitrust-specific concerns such as market definition besides the point when it comes to drug pricing challenges?
7. Random questions
* Can imposing AT liability really hold down aggregate drug costs? That is, if we accept the industry’s (oft maligned) claim that high drug prices are necessary to subsidize all the research into drugs that never pan out, then if we make it harder for firms to maintain monopoly pricing on marginal drug patents, won’t they just be forced to raise prices on other drugs with strong patents to recoup the losses (or, if that’s not possible, to scale back R&D)? In other words, are the $35 billion savings proposed by FTC illusory?
*Many commentators assert either (1) that it’s desirable to allow early entry of a generic drug even where the branded patent would have been deemed valid if litigated or (2) that where a patent would have been upheld, early generic entry has no pro-competitive effect. Is either position defensible? Do we want to encourage nuisance suits which may have the impact of degrading valid patents? Alternately, even if a patent is valid, isn’t replacing patent monopoly with generic competition by definition pro-competitive (even if legally unwarranted?) On which side should we err?
* Is FTC’s $35 billion analysis conceptually sound? I’m thinking here not so much of the numeric assumptions FTC makes, but more in terms of whether they are ignoring countervailing factors. For example, generic drugmakers argue that a per se ban on PFD would result in fewer ANDA challenges being brought against branded drugs, so that some patents which currently fail would maintain monopoly status under a ban. Is this a realistic concern, and if so how should it be weighed in the equation?
* As with GBS, there’s a question of the proper but-for world for AT comparison: should a PFD settlement be evaluated against a pure early entry agreement (without reverse payments) or should we assume that absent PFD there may be no settlement, in which case the but-for world is either (1) no challenge at all (monopoly pricing for the full term of the patent) or (2) litigation, with a total win for one side or the other.?
Chillin'Competition 
Paul, in response to question 6 raised above, I tend to think that a more thorough patent validity review at the outset, whether that process be adversarial or something else, would do much to alleviate this problem. One of the major issues, it seems, with this process is that both sides are making a guess based on their own subjective (and untested) measures of patent strength. When these guesses are off-base, as has happened a few times in recent history, strange things happen. Obviously a change in the patenting procedure for these drugs would totally overhaul the current Hatch-Waxman system (for example, there’d no longer be first filers or anything like that, as the patent would never be granted in the first place), but it might also affect research and development–which seems to be the major risk to consumers that lurks at the extreme end of all of these potential changes to the system.
Kevin Fitzgerald
16 April 2010 at 5:21 pm
1. As the outset, I agree with Kevin’s comment in that procedure to obtain patent should be reviewed more carefully. If generic (surprisingly often) wins cases regarding validity of patent, we should be more skeptical to protect patent holders whose patents are litigated to be invalid. In this sense, I am not very convinced by
the following argument that Paul pointed out.
If patents are presumed valid, then patent holders should have the choice to litigate to protect their patents or settle to avoid litigation hassles, as long as the monopoly is not extended beyond the scope of the patent.
That being said, is it easy to tighten patent validity review? My guess is that patent validity review is loose to give some competitive advantage to the US pharmaeutical firms. Maybe as one of pro-patent policy, the US might have been generous to grant patent. Given that the US firms are leading the pharmaceutical industry, it might be wiser to grant patent easily in order to increase the US’s GDP. If the US is not leading pharmaceutical industry, it might have been easier to tighten patent validity procedure.
Check out figure 3 to see that the US firms are leading the industry.
http://www.duke.edu/web/soc142/team2/firms.html
If patent validity review becomes loose and it becomes easier for generic to enter the market, it might be firms in other countries (e.g., Brazil), not the US companies that increase the revenue. This is not legal but policy argument, but I thought it worthwhile considering.
2. Second, I cannot present a convincing example of pro-competitive effect of PFD deal. I have not seen such examples yet. Has anyone seen one?
Many supporters tries to justify PFD deal by legal doctrinal reason; Patent holders, even if litigated, should be entitled to have a choice to make a deal. It is difficult to determine to what extent patent holders whose validity of patens are litigated
should be entitled to extract monopoly profit from the market. It would be unacceptable to allow firms whose patents are very likely to be invalid to continue to earn monopoly profit. On the other hand, we would not like the situation where firms whose patents are very likely to be valid but want to avoid litigation (due to transaction cost) are not entitled to enjoy the benefit of patents. It is difficult to draw the line, and I do not have a clear answer for this…..
Takashige Yamada
16 April 2010 at 5:22 pm
Thanks to Kev for the shoutout in his post (-;
In response to Takeshige’s post, here’s my initial reaction based on what I’ve been reading, but I have zero expertise in this beyond what I’ve been reading this week and may be way off base.
First, in terms of making patent review more stringent up-front, my understanding is that all patents, pharma and otherwise, follow this same pattern, where a patent can be easily obtained and then easily challenged. I agree this might be to benefit US interests, but it also may just reflect that (1) few patents will be challenged (or even used) so it may be a waste of admin resources to look deeply into each application, esp. where no adverse party is present to develop an effective challenge (2) these are by defn tech- and research-heavy fields, where detailed exploration by PTO officials of whether a patent is new and obvious would be onerous.
In the specific case of pharma, I wonder whether generic mfrs would have the same opportunity and incentive to challenge a patent prior to the FDA approval process as they do post-commercialization, and whether the delay in bringing lifesaving drugs to market would be deemed worthwhile?
Second, in terms of whether any PFDs are pro-competitive, I think it’s helpful to sort out three kinds of agreements:
1. PFD where a generic does not enter the market til patent expiration; hard to justify in my mind.
2. PFD where generic does come on the market pre-expiration, but later than it would absent the payment. The argument that gets made here is that these small generic companies need the financial resources to commercialize their drugs, and so $5 mil and 3 years on the market, for example, may be more valuable to them (and by extension to consumers) than no cash and 4 years on the market. I’m not in much of a position to assess this claim, but it seems to rely on the assumptions that (1) generics dont have other sources of funding (venture capital, etc) and (2) that they are somehow entitled to $ from the branded company. Both of these seem rather dubious, so I’m inclined to agree with you here too.
3. Then there are PFDs that include related provisions. Much of the debate is here, since the stricter versions of the PFD bans floating around in Congress bans the branded company from giving the generic “anything of value” in exchange for a delayed entry. So the argument gets made that where generic delays entry by a few years, but enters into a licensing or contract manufacturing or ingredient supply agreement with branded, this may be pro-competitive since it allows generic to profitably ramp up its mfg infrastructure in preparation for broad-scale commercialization of its own drug. That’s just one of many scenarios that get proposed as being pro-competitive. In part it comes down to how we define pay for delay and how strictly one wants to read the language in these bills.
Paul B.
16 April 2010 at 5:23 pm
This is a great post. I feel like I’ve learned allot just by reading through it. So Paul, here’s some of my initial thoughts related to Question 2.
I thought it might be helpful for us if we traced how the settlement would play out from the parties’ perspectives.
* A is the brand company who has a patent for P-B. P-B is the modern day equivalent for Aspirin and works particularly well for migraines caused by difficult antitrust issues.
* B is an upstart who has developed a generic version of P-B. It sues to invalidate A’s patent.
* The outcome of the patent validity issue is unclear, and the parties wish to settle.
o There are 2 basic levers of negotiation. The cash payment from A to B (“Cash”), and the time of entry (“Entry”).
o B can recover its costs and make a profit either by accepting Cash, or entering the market. Either it is more Cash, or earlier Entry. B tells A that it is willing to sling-it-out, and that if A wants to settle, it should put a proposal on the table.
o A huddles with its lawyers to come up with a settlement proposal.
+ It first calculates the total demand for P-B. This comes to 1 million doses per annum. It then makes a projection of B’s market share over the remainder of the patent (3 yrs). Based on past performances of generic drugs in similar cases it averages out to 60% for the next 3 yrs. It also makes a reasonable prediction of the price of the generic version. This turns out to be $10 per dose. A further projects that the associated costs for producing and marketing the generic version would be $5 per dose, so B would be making $5 per dose. B’s expected profits for 3 yrs is 600K x $5 x 3 = $9 million.
+ But there are risks in entering the market, which can come from a variety of factors such as abrupt changes in the economy (e.g., think about the financial crisis) or changes in drug regulation that impact the marketability of the drug. A reasonably believes the chances of such risk altogether is 40%.
+ Based on the above, A believes B’s expectations for profits upon entering the market is $9 million x 60% = $5.4 million. This should be further discounted for the chance that B might not prevail in the patent suit, which A reasonably believes has a 50% probability. So B’s (average) expected profit upon continuing litigation is $2.7 million.
o A approaches B and shows it the number. It then explains that B can avoid the risks of entry (40%) by delaying entry for one or more years. It proposes $3 million for not entering into the market for the next 3 years. It argues that this is better than the $2.7 million they can expect by slinging-it-out in the litigation.
o B mulls over A’s proposal. It decides it wants to have both ways. It comes back with a proposal saying that it will delay entry for 2 years upon a payment of $3 million. It explains that it is willing to accept a non-market risk discounted amount of $3 million for the first 2 years ($3 million x 2 x 50%).
* The parties negotiate and settle on the amount of $2.9 million for delayed entry of 2 years. Consumers will see the cheaper generic version after 2 years.
This is a quite crude hypothetical, and the settlement in the real world might play out differently. But I believe we can make a few observations.
* A settlement that combines cash payment and delayed market entry may not simply be a devised scheme to distort competition. It may be the result of the use of 2 available levers of negotiation (Cash and Entry) that are useful due to their reliability and visibility.
o Because of the utility of both levers in settlement, there might be problems with presumptively condemning settlements that involve delayed entry for cash payments which exceed only “litigation” costs or other definitive categories. Any such bright-line rule might prove to be too restrictive (or too expansive).
* There are variables over which reasonable minds can differ, in particular (i) market risks (“X”), and (ii) probability of the patent suit outcome (“Y”). This also means that they can be subject to manipulation by the parties and can serve as a cover for sharing of monopoly rents.
o If our main concern were cases where payments were made to prolong the monopoly of “questionable” (weak) patents, one might suggest a rule where the parties could not set Y above 50%. This would mean that only patents with more than a 50% chance of maintaining its validity would be subject to PDF settlements. But, this might then facilitate more manipulation of X, or other variables in the equation. Also, such numbers are so subjective that it would be extremely difficult to monitor and police such a rule.
Because of the multitude of ways the parties could explain the rationale for the settlement, there seems to be a non-insignificant chance of the DOJ’s approach resulting in a near per se legal rule depending on its application. In this regard, there seems to be more of a difference between the approach of the DOJ and that of the FTC (presumptively illegal approach) than what some of our readings suggested.
I think that, to some extent, the issue boils down to how you weight the 2 following considerations:
* “More tangible and immediate, but perhaps more short-term” consumer benefits flowing from generic entry and lower prices as a result of proscribing PDF settlements.
* “Somewhat indirect and perhaps questionable, but more long-term and possibly more substantial if realized” consumers benefits flowing from more innovation as a result of more vigorous patent protection.
Final note: There could be an argument that innovation would suffer because generic companies would have less of an incentive to develop generic drugs. Well, a few questions come to mind. (i) Is protecting the opportunity for a rush to be the first to get paid-off the right way to foster innovation, and does it provide the right incentives? One might worry that what we would end up with is innovation, but just enough to get paid-off by a branded rival, rather than that enough to get the product succeed in the real market (or at least at an earlier time). (ii) If some generic competitors decide to drop out of the race due to diminished expectations, and only the hearty few (that believe they can win the law suit on the merits, or make more money by earlier Entry rather than receiving more Cash) remain, is that really a tragic outcome? Unless, of course, it’s likely that nobody would race at all.
Yong Lim
16 April 2010 at 5:24 pm
>Paul
Thanks for your response.
(1) few patents will be challenged (or even used) so it may be a waste of admin resources to look deeply into each application, esp. where no adverse party is present to develop an effective challenge (2) these are by defn tech- and research-heavy fields, where detailed exploration by PTO officials of whether a patent is new and obvious would be onerous.
=>I see. would granting IP too loosely hinder innovation (to disincentivize competitors and other firms that wish to use the relevant technology)? In this regard, I am still a bit skeptical whether we can say more careful patent review is
“onerous” or not. Nonetheless, I am far from patent expert, so this is only my intuiition..
>Yong
Very interesting comment! I have a question. I understand how settlement works out and parties negotiate settlement by reflecting preference regarding various facgtors such as cash, entry, market risk and probability of the patent suit
outcome. However, does the fact that parties consider various factors and made a deal could justify the consequence that brand firm and generic firm share monopoly profit? It seems to me that, if it cannot be justified by IP, it is difficult to allow these two parties to share profits (I might be misleading your argument….if it is a case, please correct me.).
Second, I agree with you that we have to balance the short-term benefit(low price as a result of generic entry) and long-term benefit (incentive to further R&D, especially for brand firms). But would it be appropriate to take long-term benefit when we consider the PDF deal? Congress attemps to respect this part in the framework of Patent Law and, probably, Hatch-Waxman Act, right? If so, Antitrust should respect duly exercise of IP law. In this regard, we can frame question such as “whether or not PFD deal is within the scope of Patent law and Hatch-Waxman Act”? Things are getting difficult and I think that I have migraines, so I need P-B…
Takashige Yamada
16 April 2010 at 5:25 pm