Skinny Label Safe Harbor

by Dennis Crouch

Most of the Orange Book patents do not cover a novel active ingredient or new drug. Rather, most are new methods of treatment or formulations.  When a drug is covered only by method-of-treatment claims, that means that generic sales of the active ingredients does constitute patent infringement, unless the sales are somehow inducing folks to perform the method.  This gets a bit complicated when a drug has  multiple therapeutic uses, only some of which are patented.  Obviously, generic versions should be permitted for the non-patented uses.  But, we know that there is going to be a lot of substitution at the retail level because of the price differential between the brand & generic.

Skinny Label: Congress thought through this problem (a little bit).  The basic approach is that the Brand (Patent Owner) must tell the FDA (under penalty of perjury) which of its labelled & approved uses are protected by patent.  When approving a generic version, the FDA will respect the patent by carving-out the patented uses for the generic’s “skinny label.”  Of course, doctors, pharmacists, and patients all know that the generic version is the same drug at a lower price and so go ahead and substitute the generic version even for the patented uses.  In these cases we’re talking about many millions of dollars in sales. That means we have very sophisticated parties on all sides–everyone knows what is happening.

The statute does not create any express ‘safe harbor’ for the generic manufacturer against liability for patent infringement.  Certainly, the manufacturer would be liable for inducement if it encouraged off-label infringing use of its product.  But, does the manufacturer need to take a more active role in discouraging infringing use?

This situation was litigated most recently in GlaxoSmithKline LLC v. Teva Pharm. USA, Inc., 7 F.4th 1320 (Fed. Cir. 2021), en banc denied at GlaxoSmithKline LLC v. Teva Pharm. USA, Inc., 25 F.4th 949 (Fed. Cir. 2022).  GSK sells the beta blocker carvedilol labelled for three uses:

  1. treatment of congestive heart failure
  2. treatment of hypertension
  3. treatment of left ventricular dysfunction following myocardial infarction

All three uses were patented, but by 2007, patents covering the second and third uses had expired.  At that point, Teva received approval for its generic version, but with a skinny label that carved-out the congestive heart failure usage.  The procedural posture of the case is somewhat confusing, but the Federal Circuit eventually sided with the patentee–holding that Teva’s skinny-label along with other practices induced infringement (i.e., encouraged folks to use it for congestive heart failure).

In a recent filing, Teva’s attorney William Jay (Goodwin Proctor) indicated that Teva plans to petition the Supreme Court for writ of certiorari in the case.   (Request for Extension). If Teva’s request for extension is granted, the petition will be due July 11, 2022 with any amicus in support due in early August.  Teva’s failed en banc petition provides a guide for questions that it will likely ask in its petition:

(1) Can the generic manufacturer nonetheless be held liable for induced infringement based on evidence that would be available in every carve-out case—the skinny label itself and product materials that describe the generic drug product as the AB-rated generic equivalent of the brand product, but do not even mention the patented method?

(2) Even if the generic manufacturer were found to have encouraged infringement, can it be held liable for infringement that it did not cause—e.g., if the direct infringer undisputedly did not see the communication that supposedly encouraged infringement?

Teva En Banc Petition.


20 thoughts on “Skinny Label Safe Harbor

  1. 4

    Jansen v. Rexall Sundown, Inc., 342 F.3d 1329 comes to mind re labeling. It seems many parallels present.

    1. 4.1

      Interesting case – so thanks for that.

      But the parallels seem to fall far short of the FDA-imposed labeling.


          The Jansen case is about a company’s decisions on labelling – there is no intersection with forced labelling (skinny or otherwise) from the FDA.


            Right on, and thx. But gov can’t force a merchant to induce others to infringe a patent, to the detriment of the merchant, can it ? There’s been questionable (in my mind anyway) decisions about what constitutes inducement. Is it a question of law ? Or should we be on a fact-finding mission, interviewing persons of ordinary skill (customers), and plop the product into their lap and ask them “What does knowing what you now know, make ya wanna do ?” Maybe a properly-phrased label-license could work…. ?


              But gov can’t force a merchant to induce others to infringe a patent, to the detriment of the merchant, can it ?

              An interesting question.

              Not my bailiwick, but I think the answer is yes – if the government is not artful in its directions, I can easily see those directions leading to an induced infringement. To be clear though, I do not think that such is there in the present fact pattern.

  2. 3

    Simply require the application of a “won’t improve congestive heart failure” coating on the generic version.

    Voila! Problem solved.

    (I mean — hey — if spraying Lysol or gargling bleach will kill the Covid virus . . . )

  3. 2

    As understood, there is a potentially huge difference in U.S. drug costs impacted by FDA-approved “skinny labeling” of an unpatented drug for only one usage being treated as proof of induced patent infringement for any other usage under the Fed. Cir. Teva decision. Thus, it would amaze me if the Teva cert petition, once actually filed and publicized, does not generate several amicus briefs, perhaps even from some members of Congress, and much higher than average odds of the Sup. Ct. granting cert. But I would be interested to see counter-arguments?

    1. 2.1

      It is said that it is “better to be lucky than good.” GSK was both lucky and good here. They were lucky that the language that they needed to include on the label for safety purposes had the oblique effect of suggesting an on-patent use, and they were good (or at least their patent prep&pros team were good) in that they crafted patent claims that spoke to that label language.

      This is an idiosyncratic situation that would be hard to replicate more generally. Teva has a vested interest in making out as if this case represents the death of the skinny label, but in the time since this decision came down, the CAFC has adjudicated other cases in which the generic won on a skinny label defense. The rumors of the death of the skinny label defense are much exaggerated.

      I hope that the SCOTUS does not take cert on this one. If they are interested merely because they think that the CAFC’s decision could blow up drug costs, that would be a regrettable mistake.

      1. 2.1.1

        Greg, “an idiosyncratic situation” is the opposite of what the Fed. Cir. dissenters, and prior amici, to both the first and “revised” Fed. Cir. opinions, are quoted as saying in the above-cited cert petition time extension request.


          I would be curious as to these additional decisions.

          If they are anything like 101, a new panel decision having a different takeaway may merely be indicative of a different problem and may not in fact give Greg his position of “non-death of the skinny label.”

          As this is not my bailiwick, I am just familiar with these additional decisions.

          Further, if these additional decisions are NOT distinguishable, then these actually MAKE good cases to change the result of the instant decision.


          As I said, Teva’s interest is in making out that their loss is a society-wide disaster and not a mere quirk. Of course they will quote those saying as much. For whatever that is worth…


          To put it another way, is this alleged Fed. Cir. presumption of induced infringement from an FDA approved label being accused of creating a de facto monopoly over the sale of an unpatented drug by a patent on only one use of the unpatented drug? If so, that kind of briefed accusation should get more than usual cert petition review.


            The way that you put that, I would expect less rather than more. Your restatement appears to provide a cut-n-dried situation for use simply does not need Supreme Court review.

  4. 1

    “disparate portions of Teva’s skinny label could be patched together to match up with all the limitations of GSK’s CHF method patent”

    My initial reaction is don’t do that. Not my area of expertise, so is every word on the label mandatory? Does the FDA forbid CYA disclaimers (“…not recommended for treatment of congestive heart failure…”)?

    1. 1.1

      The FDA’s single most imperative concern on drug labels is avoiding confusion at the point of care. They would never let a generic include “not recommended for” CYA language if the branded version of the same drug has been shown safe and effective for a given use. They would not want anyone left with the impression that active drug ingredient X is not “recommended” for a given usage if it actually is recommended for that usage in the branded context.

      1. 1.1.1

        Perhaps then a label similar to that used by non-pharm industry on the order of “Not to be sold individually,” but rather, “To be sold ONLY for…”

Comments are closed.