The interaction of the patent exhaustion doctrine and covenants not to sue was highlighted in a recent opinion from the District of Delaware in Purdue v. Collegium. The court denied Collegium’s motion to dismiss, which was premised on Purdue’s covenant not to sue Collegium’s supplier.
Purdue sued Collegium for infringing U.S. Patent Nos. 9,861,583; 9,867,784; and 9,872,836 with its pain-relief medication Nucynta. Separately, Purdue settled a patent-infringement suit with Assertio, another pharmaceutical company, which involved granting Assertio a covenant not to sue. The covenant not to sue extended to other entities interacting with Assertio, but Collegium was expressly carved out.
The settlement affected this case because Assertio supplied Nucynta to Collegium under a commercialization agreement. That commercialization agreement provided that title to the product shifted from a contract manufacturing organization (CMO) directly to Collegium, never passing to Assertio. But the day after Assertio received the covenant not to sue as part of settling with Purdue, Collegium and Assertio amended the commercialization agreement so that title did transfer from the CMO to Assertio before transferring to Collegium.
The court’s reasoning was based on interpreting this commercialization agreement against the background of the patent exhaustion doctrine. Generally, an authorized sale of a product triggers patent exhaustion: when the patentholder (or a third party licensed or otherwise given permission by the patentholder) sells a product to a customer, the customer does not infringe the patent by using or selling that product. Sales pursuant to a covenant not to sue qualify as “authorized sales” that trigger patent exhaustion.
With that background, the court interpreted the commercialization agreement under two scenarios: ongoing sales starting after the amendment, and sales made before the amendment. The court found that ongoing sales made as described in the amended commercialization agreement—that is, with title transferring to Assertio and then to Collegium—would trigger patent exhaustion, saving those products from infringement.
Under the second scenario, the court noted disagreement between district courts about whether covenants not to sue were retroactive with respect to patent exhaustion. The terms of the covenant not to sue did not resolve the issue because the covenant did not explicitly state whether past sales were covered. The court left this issue unresolved.
The court ended up not needing to resolve the patent exhaustion issue with respect to either scenario because it was unclear whether the facts really matched what was described in the commercialization agreement. Both pre- and post-amendment, the commercialization agreement stated that the title transfer was subject to supply agreements with the CMOs. Those supply agreements were not in the record, so the court could not decide whether or not those agreements changed how title was transferring. As a result, the court denied the motion to dismiss; however, the door is still open for Collegium to win on this issue.
Lessons for Practice
The key takeaway is the legal point that a covenant not to sue can cause patent exhaustion. The recipient of a covenant not to sue becomes a source of necessarily noninfringing competing products, absent the terms of the covenant not to sue stating otherwise. A patentholder needs to consider the possible downstream effects of a covenant not to sue.
Trying to manage those downstream effects requires considering all the possible routes by which noninfringing products could move through the covenantee. If the patentholder wants to carve out a third party, that carveout should explicitly spell what is not authorized, for example, selling to the third party, sublicensing to the third party, selling directly to customers of the third party with a finder’s fee going to the third party, and so on.