It was a stunning number, purportedly the largest patent settlement in history. During trial, defendants Teva ($1.6 billion) and Sun ($550 million) agreed to pay plaintiffs Wyeth and Nycomed a combined $2.15 billion to settle litigation stemming from their “at-risk” launches of generic versions of Wyeth and Nycomed’s Protonix product.
At-risk launches have become a popular tactic by generic companies over the last 10 years. As a result, counsel for branded and generic companies are increasingly litigating highly complex damages issues. However, the law on patent damages as applied to the interactions between generic and branded products in the pharmaceutical marketplace remains largely undeveloped.
Before 2002, the idea that a generic company would begin selling its product before a favorable decision from the U.S. Court of Appeals for the Federal Circuit seemed only a theoretical possibility. The risk that a product launch would result in damages exceeding its profits, together with the purpose of the Hatch-Waxman Act itself, appeared to mean that no generic company would launch before such certainty (generic drugs are typically sold at a substantial discount to the brand). That changed in June 2002, when Geneva Pharmaceuticals began selling a generic version of the popular antibiotic Augmentin after winning summary judgment of patent invalidity. Geneva’s decision provided a new paradigm for generic companies, one that quickly morphed. Soon, generic companies were no longer waiting for even a district court victory to begin selling. Rather, they were actively positioning their cases to be in a position to launch at the end of the Hatch-Waxman imposed “30-month stay” on FDA’s approval of a generic drug. Aided by the Federal Circuit’s shifting standard for preliminary injunction,1 the threat was reinforced in settlement negotiations. Faced with the potential sudden loss of a large chunk of revenue, many cases were settled. But, at least 26 times since Geneva’s actions, generic companies have followed through on the threat. Until Protonix, only three had gone to trial, and none yielded any significant case law guidance for some of the unique issues presented by patent damages in the pharmaceutical marketplace. Settlement likely disappointed those seeking more clarity on potential measures of patent damages, including lost profits and reasonable royalty.
To recover lost profits, a patentee must show it would have received that profit “but for” the infringement.2 The lost profits analysis typically includes four Panduit factors: proof of demand for the patented product or feature, an absence of acceptable non-infringing alternatives meeting that demand, the ability of the patentee to satisfy the demand, and the quantum of lost profits.3 Showing demand for the patented product seems not to be an issue, to date, in such cases. After all, the branded product’s sales show that demand and probably the ability to satisfy demand as well. However, any history of manufacturing issues affecting the ability to supply the market will likely be cited as evidence of an inability to meet demand.
Determining if there are acceptable non-infringing alternatives first requires defining the marketplace. No case law answers that question definitively for this unique market. The patentee will likely seek to confine the market to that branded product itself. It will argue that the generic copy is a direct substitute for the brand, and that this defines the market—when a prescription is written for Prozac, the pharmacist cannot substitute another anti-depressant absent the doctor’s permission. Because most states mandate generic substitution, generics can quickly capture up to 90 percent of branded sales. The patentee will also point to the defendant’s business model, a model based on substitution for specific brands. Finally, the patentee will cite the Hatch-Waxman statute, and argue that there can be no non-infringing alternatives under its framework. On the other hand, the generic defendant will argue for a broader marketplace, in which, by definition, non-infringing alternatives exist. Thus, for a drug like Protonix, defendants may argue that the marketplace is all proton pump inhibitors, or even all anti-ulcer medicines. While the presence of non-infringing alternatives does not completely deny the availability of lost profits, it can consign the patentee to a damages award based on market share.4 Under a market-share theory, a product with 20 percent market share would capture 20 percent of the generic’s sales. The plaintiff might thus receive a lost profits award limited to only 20 percent of the infringing sales—even though the generic captured 90 percent of branded sales. Existing law has not yet addressed the equities of this scenario.
Branded companies sometimes respond to an at-risk launch by selling an “authorized generic” version of the drug, competing with the infringing generic product on somewhat even grounds, and mitigating loss to some degree. But the generic defendant may argue that the patentee’s authorized generic version of the drug is a non-infringing substitute. The grant of a patent license to a third party, which then sells a product covered by the same patent, can create a non-infringing alternative in the marketplace.5 It is unknown, however, whether the Federal Circuit would apply that case law to this particular market. A defendant may also argue that the “authorized generic,” although launched in response to the defendant’s actions, has accelerated the loss of brand sales, and that it should not be responsible for such self-inflicted harms. The response by the patentee: “But for your infringement, I would not have launched my own generic product, and I would have continued to sell at least at prior levels and prices.”
A patentee must also prove the quantum of profit it would have received “but for” the infringing sales, typically relying on actual historical sales and forecasts of future sales for this proof. Of course, the historical accuracy of the forecasts will be contested. But the issues go far beyond. For example, an infringer may point to factors outside of the brand-generic interplay, attempting to place blame for loss of sales elsewhere. It might argue that sales would have declined anyway because a competitor brand (or a generic to that brand) has already been capturing sales. Or a defendant may argue that at least some of its sales represent an expansion of the market that the brand would therefore not have made, and thus are sales for which lost profits damages are not available. Little guidance from the courts addresses these particular issues directly.
A patentee may also argue that it is entitled to lost profits damages based on price erosion. Yet the sales price of a branded product often facially increases following generic entry. That, however, does not account for rebates and/or discounts given to customers. This may require a customer-by-customer analysis. Where the patentee responds to a generic threat by selling an authorized generic, the patentee may also seek price erosion damages for that product. But the generic may argue that as a separate product no price erosion damages are available. The generic may also argue that authorized generic sales actually caused price erosion—the entry of multiple generic products can cause a steeper price erosion curve than a single entrant—and that, at a minimum, they should not be held responsible for such additional erosion. No case law definitively provides an answer.
Where lost profits damages are not available, a patentee is entitled to receive a “reasonable royalty” as compensation for infringement. On its face, the formula is simple. Determine the royalty base—number of infringing units sold (for which lost profits are not available) and sales price, i.e., sales revenue—and multiply by the royalty rate. But determining the royalty rate is not so simple: What royalty rate would a willing licensor and licensee agree to in a hypothetical negotiation held at the time infringement began. This hypothetical analysis is informed by as many as 15 Georgia-Pacific factors.6
Because of the uniqueness of the Hatch-Waxman law, however, it is unclear when infringement can be said to begin—and thus the date on which the hypothetical negotiation occurs. Under that statute, the filing of an application with the FDA to market a generic version of a branded drug with a certain type of patent certification is a technical act of patent infringement designed to vest the courts with jurisdiction.7 Thus, the hypothetical negotiation could occur that early. But because of the statutory 30-month stay on FDA approval, actual commercial infringement begins several years thereafter. The hypothetical negotiation date matters, at least in part, because relative bargaining power can shift over time as between the parties. To date, there is no conclusive answer on when infringement begins for this purpose.
One of the most important Georgia-Pacific factors is a party’s own licensing history for the same patents or comparable technology/patents. Such licenses can be used as evidence of a party’s willingness to license its patents and the rates it demands or is willing to pay for such technology. The Federal Circuit has taken a keen interest in comparability of technology and license agreements,8 and experts routinely battle over which licenses are “comparable.” Hatch-Waxman cases are no different. Consider an extended release drug for the treatment of type-2 diabetes containing drug X as its active ingredient. Comparable technology or patents could include at least those for any product containing drug X; those for extended release drugs generally; those for the treatment of type-2 diabetes; or subsets or combinations of each. Disputes regarding the comparability of technology often first arise during discovery, leading to costly litigation maneuvering in an attempt to gain, or prevent disclosure of, license agreements.
Patent licenses between innovator pharmaceutical and generic companies are generally uncommon—except in the context of settling patent litigation. The Federal Circuit has held, though not in the context of Hatch-Waxman litigation, that patent licenses entered into in settlement of litigation can be the most relevant evidence in appropriate cases.9 Parties to Hatch-Waxman litigation often dispute discoverability of such agreements, as well as their probative value as evidence of a royalty rate to which a willing licensor and licensee would agree. Unlike many patent licenses, the innovator-brand license agreement is typically not driven by royalty rate. Rather, the date on which a generic company may begin selling its product is often the most significant factor during negotiations. Value for the branded plaintiff is driven by the length of time during which it enjoys exclusivity without generic competition. Parties bargain hard over this date, the generic seeking the earliest possible entry date, the patentee the latest. Therefore, settlement agreements in Hatch-Waxman litigation may not fairly represent an appropriate royalty rate for these purposes. Yet, generic defendants will continue to point to licenses bearing low royalty rates and patentees will debate the probative value of such agreements. Alternatively, a patentee may argue that the time element to such licenses is where the value truly lies, and attempt to place a monetary value on that time element. In any event, there is little case law directly addressing these particular issues.
In addition to lost profits and/or reasonable royalty damages, a patentee may also be entitled to other consequential damages. Where a generic company launches at-risk, the branded pharmaceutical manufacturer may be forced to close facilities and lay off employees. It may lose discounts for bulk purchases of raw materials, suffer from reduced production efficiencies, bear increased financing-related costs and lose investment opportunities due to a reduction in revenues resulting from generic sales. Damages for similar harms have been affirmed by the Federal Circuit in other contexts.10 Whether the branded-generic context poses a unique fact scenario rendering such precedent inapplicable remains an open question.
Assuming the generic defendant is enjoined from further infringing sales, the patentee may also continue to suffer harm from the at-risk launch long after the infringing product is removed from the market. For example, it may be unable to recover its former sales levels, market share, or pricing levels. Damages for such harms are recoverable if the patentee can prove future economic harm by providing sound economic proof of the nature of the market and likely outcomes, with infringement factored out of the economic picture.11 Speculative recoveries are not permitted. Undoubtedly, the generic defendant will argue that with its product removed from the market, the brand will fully recapture prior sales levels, and profitability. The patentee, however, may provide evidence showing that the generic entry has permanently altered the marketplace. The branded manufacturer may have been forced to increase its discounts and rebates to maintain drug formulary coverage or placement. It is unlikely to be able to retreat from such discounts after the generic product is removed from the market. Or the branded product may have been placed in a less preferred “tier” by one or more formularies, as a result of the generic entry. If so, the brand may find it difficult to recapture its prior position. In any event, claims for future damages are sure to be hotly contested, again, with a lack of precedent specific to the industry.
Those within the pharmaceutical industry had hoped that the Protonix litigation might provide some guidance to those litigating in this area. And maybe it did, even without Federal Circuit review.12 At trial, the plaintiffs informed the jury they were seeking about $2.7 billion in damages from the two defendants ($1.9 billion from Teva, $838 million from Sun), as a combination of lost profits and reasonable royalty damages. The settlement with Teva ($1.6 billion) represented only about a 16 percent discount from the amount plaintiff sought, perhaps reflecting the relative strengths of the parties’ legal arguments.
David Manspeizer is a partner at Wilmer Cutler Pickering Hale and Dorr, and the former vice-president, Intellectual Property, at Wyeth.
1. See e.g., Abbott Labs. v. Sandoz, 544 F.3d 1341 (Fed. Cir. 2008).
2. See, e.g., King Instruments v. Perrigo, 65 F.3d 941 (Fed. Cir. 1995).
3. See Panduit v. Stahlin Bros. Fibre Works, 575 F.2d 1152 (6th Cir. 1978).
4. See, e.g., State Indus. v. Mor-Flo Indus., 883 F.2d 1573 (Fed. Cir. 1989).
5. See, e.g., Pall v. Micron Separations, 66 F.3d 1211 (Fed. Cir. 1995).
6. Georgia-Pacific v. U.S. Plywood-Champion Papers, 318 F. Supp. 1116 (S.D.N.Y. 1970), modified sub nom. Georgia-Pac. v. U.S. Plywood-Champion Papers, 446 F.2d 295 (2d Cir. 1971).
7. See 35 U.S.C. §271(e)(2)(A).
8. See, e.g., ResQNet.com v. Lansa, 594 F.3d 860 (Fed. Cir. 2010).
9. See, e.g., ResQNet.com, 594 F.3d at 869-72. No settlement negotiation privilege applies to “settlement negotiations related to reasonable royalties and damage calculations.” In re MSTG, 675 F.3d 1377, 1348 (Fed. Cir. 2012).
10. See, e.g., Mach. v. Magna-Graphics, 745 F.2d 11, 22 (Fed. Cir. 1984) (award compensating patentee for its decreasing marginal cost of producing goods); LAM v. Johns-Manville, 718 F.2d 1056, 1065 (Fed. Cir. 1983) (award compensating for increased promotional expenses).
11. Grain Processing v. Am. Maize-Prods., 185 F.3d 1341, 1350 (Fed. Cir. 1999); see also Oiness v. Walgreen, 88 F.3d 1025 (Fed. Cir. 1996); Water Technologies v. Calco, 850 F.2d 660 (Fed. Cir. 1988); LAM, 718 F.2d at 1065 (Fed. Cir. 1983).
12. Infringement and validity had already been decided by another jury in 2010. Appeal on those issues presumably also awaited judgment on damages.