Adams Respiratory Therapeutics, Inc. v. Perrigo Company (Fed. Cir., August 5, 2010)
Panel: Linn, Moore (author), Friedman
By Jason Rantanen
Adams holds patent number 5,372,252, which covers an extended release formulation containing guaifenesin (an expectorant used to thin, loosen, and help expel mucus that causes congestion). Perigo sought FDA approval for a generic version of Adams' product, Mucinex®. After construing the claims, the district court granted summary judgment of noninfringement. Adams appealed that decision.
The claim terms in dispute related to pharmacokinetic parameters. These parameters are used to characterize the rate and extent of absorption of the active pharmaceutical ingredient ("API"). The primary term at issue, Cmax, indicates the maximum concentration of the API following dosing.
"Equivalent Cmax"
The parties' main dispute was over the meaning of the term "equivalent" in the context of "provides a Cmax in a human subject equivalent to the Cmax obtained when the first of three doses of a standard immediate release formulation having one third the amount of guaifenesin is dosed." For purposes of FDA approval, a formulation is considered to be bioequivalent if, among other things, its Cmax is within 80% to 125% of the value with which it is being compared at a 90% confidence interval. The district court agreed with Perrigo that "equivalent" should be construed to mean equivalent under the FDA bioequivalence guidelines.
On appeal, Adams challenged the requirement of a 90% confidence interval, arguing that it makes sense when seeking FDA approval, but not when proving infringement.
The Federal Circuit agreed with Adams. In doing so, it rejected Perrigo's argument that the inventors had expressly defined "equivalent" as te FDA's bioequivalence guidelines. Rather, the court construed Adams' reference to the FDA guidelines as referring specifically to the 80-125% range, not to the requirement of a 90% confidence interval. According to the court:
Requiring a 90% confidence interval would inappropriately raise the bar for establishing infringement. Adams must show that it is more likely than not that Perrigo's ANDA product will have a Cmax within the 980-125% range. Adams is not required to show that Perrigo's product will meet this requirement 9 times out of 10.
Slip Op. at 8.
Comment: This interpretation opens up a box of statistical worms, and I suspect that down the line the court may regret its venture into probability theory. For the time being, however, it gives parties useful language to quote when attempting to prove infringement (and perhaps invalidity) via clinical results.
A ≈ B ≈ C therefore A ≈ C
Adams also appealed the district court's ruling that it impermissibly compared the accused product to Mucinex. Adams' argument was that because the accused product was bioequivalent to Mucinex, and Mucinex was bioequivalent to a standard immediate release ("IR") product, then the accused product had a Cmax equivalent to the IR product.
The Federal Circuit agreed that, under the circumstances of this case, Adams' argument was sufficient to preclude summary judgment of noninfringement. The court cautioned, however, that "[i]f Adams had relied on the mere fact of bioequivalence of the two sets of products (and no PK data or Cmax values, that would not be enough to survive summary judgment." Slip Op. at 11. Here, however, Adams presented actual PK data and Cmax values, which a fact-finder could look at when assessing equivalence between the accused product and an IR product.
"Bioavailable"
The court also addressed the meaning of the term "bioavailable" in the context of the '252 patent. The dispute hinged on whether the phrase "fully bioavailable in the subject's stomach" meant "both release and availability in the stomach for absorption, wherever that absorption might occur."
Perrigo argued that because "bioavailable" is commonly understood to mean absorption, thus requiring the guaifenesin to be absorbed in the stomach. Because guaifenesin is primarily absorbed in the small intestine, this construction would preclude a finding of infringement.
The court rejected the proposed construction as inconsistent with the specification: "Although the specification never expressly defines bioavailable, it uses the term when describing the availability of the drug for absorption, not the actual absorption." Slip Op. at 14. The court further noted that Perrigo's construction would exclude the preferred embodiment, which "is rarely, if ever, correct and would require highly persuasive evidentiary support."
Doctrine of Equivalents
Finally, Adams argued that the district court erroneously precluded it from relying on the doctrine of equivalents with respect to a dependant claim requiring that the total amount of guaifenesin released in to the patient be at least 3500 hr*ng/mL. The panel concluded that the use of a numerical limit did not preclude Adams from arguing that an amount of 3494.38 hr*ng/mL was equivalent to 3500 hr*ng/mL.
Happiness always generous person concomitant with, Riches always accompanies with faith, Wisdom always and noble person concomitant, Charm will always be with sunshine person concomitant, Let all the wishes with you always.
Type test
Type test
“A ≈ B ≈ C therefore A ≈ C”
Nope. Logically and mathematically this is a false statement. The “therefore” makes it false because the result need not follow from the first and second premise.
If approximates means, say, w/in 10% then:
A = 100
if B = 95, ergo A ≈ B
if C = 87, ergo B ≈ C but A does not ≈ C.
And I believe that is the point Moore was making. By providing the actual values, Adams was arguing A = B, B = C, therefore A = C, i.e. EQUIVALENCE, which is what bio-equivalence refers to.
6,
…according to which serum analysis? – I think ya left the claim a little indefinite.
“A modified release tablet having two portions, wherein a first portion comprises a first quantity of guaifenesin in an immediate release form which (we don’t want to tell you what form, and instead will tell you what happens to the form when it undergoes certain method steps, i.e. application to a subject’s stomach) becomes fully bioavailable in the subject’s stomach and a second portion comprises a second quantity of guaifenesin and a release-delaying matrix comprising a hydrophilic polymer and a water-insoluble polymer wherein the weight ratio of said hydrophilic polymer to said water-insoluble polymer is in the range of from about 1:1 to about 6.8:1, wherein said tablet (is of indeterminate structure but again, we’ll tell you some things that happen to the product when it undergoes some method steps) demonstrates a Cmax in a human subject equivalent to the Cmax obtained when the first of three doses of a standard immediate release formulation having one third the amount of guaifenesin is dosed every four hours over a 12 hour period, and wherein said tablet also provides therapeutically effective bioavailability for at least twelve hours after a single dose in a human subject according to serum analysis.”
That’s what the claim should have said Pharma.
Malcom, here is claim 1:
A modified release tablet having two portions, wherein a first portion comprises a first quantity of guaifenesin in an immediate release form which becomes fully bioavailable in the subject’s stomach and a second portion comprises a second quantity of guaifenesin and a release-delaying matrix comprising a hydrophilic polymer and a water-insoluble polymer wherein the weight ratio of said hydrophilic polymer to said water-insoluble polymer is in the range of from about 1:1 to about 6.8:1, wherein said tablet demonstrates a Cmax in a human subject equivalent to the Cmax obtained when the first of three doses of a standard immediate release formulation having one third the amount of guaifenesin is dosed every four hours over a 12 hour period, and wherein said tablet also provides therapeutically effective bioavailability for at least twelve hours after a single dose in a human subject according to serum analysis.
I don’t know whether the functional language was required to distinguish the prior art.
If the accused product met the physical claim limitations, maybe the functional language was all that was left to argue about.
Or, if the accused prodcut did not meet the physical claim limiations, maybe you would argue that it was “equivalent” if it still met the functional limitations.
Paul, that is the issue that caught my eye.
So should this be a matter of literal infringement, doctrine of equivalents or both of them?
At the appellate level it probably doesn’t matter which as long as they decide whether they’ve found infringement. From a theoretical standpoint, it should probably be literal if the infringer is within manufacturing tolerances or reasonable experimental error of the claimed range, by equivalents if the infringer is so close it makes no practical difference (as was found in this case), and non-infringement if the patentee intended to claim a precise endpoint.
It sounds like someone has their dimensions wrong, though. The total amount of medicament absorbed by the patient should be dimensionless in time.
The judgment contains a useful footnote:
We caution that the term 3500 hr*ng/mL should not be read “with greater precision than the claim lan-guage warrants.” Phillips, 505 F.3d at 1377. “In some scientific contexts, ‘1’ represents a less precise quantity than ‘1.0,’ and ‘1’ may encompass values such as 1.1 that ‘1.0’ may not.”
So should this be a matter of literal infringement, doctrine of equivalents or both of them? If measurement error is allowed for, it is difficult to see that there is any difference bewteen the alleged infringement and the claim.
Would anyone recommend referencing the FDA bioequivalence guidelines to define terms within a patent? Sure it would arguable narrow some terms, but in reality, if the “invention” needs to be approved by the FDA to have commercial viability, what is there to lose?
“What is this method language doing in a pharmaceutical composition claim?”
Mhmmm. There’s a riddle. This case is little more than bad advocacy on the part of the defendant’s counsel. Failure to point out how these limitations limit the claim and invalidate it over 101 or 112 is irresponsible at best.
Malcolm, as I recall, the patent claims are largely written in functional terms. It is (or was) a common practice for dosage forms.
Let’s look at the limitation: “provides a Cmax in a human subject equivalent to the Cmax obtained when the first of three doses of a standard immediate release formulation having one third the amount of guaifenesin is dosed”
What is this method language doing in a pharmaceutical composition claim? Are you telling me that the inventors can’t describe the composition in structurally descriptive terms?
I don’t have a problem that the language is in the claim. I’m just not sure why it’s being construed when it should be ignored. And the idea that there could be d.o.e. equivelents of such functional language is absurd.
Malcolm, as I recall, the patent claims are largely written in functional terms. It is (or was) a common practice for dosage forms.
I think I agree with James that there is no inherent inconsistency between the preponderance of the evidence standard and the 90% confidence interval. You could have to establish that it’s more likely than not that the accused product yields a result within the range at a 90% confidence interval.
Couldn’t the “more likely than not” requirement still apply if you accepted the 90% confidence interval?
The Plaintiff would need to show that on a balance of probabilities, the defendant’s products fall within the test range (whatever percentage and confidence level the test range calls for). This could be done by showing a) defendant’s product was tested b) by a person competent to due the test c) under proper testing conditions and d) defendant’s product failed/passed the test.
There is a difference between claims that recite a range (with or without significant figures) wherein the range refers to a structural measurement.
Here the range refers to a functional measurement. In this sense, it is more like a means-plus-function claim in which case the patentee is not entitled to equivalents of the recited function. This case is a complete eff-up by the Federal Circuit.
I wrote about another issue in this case:
link to pharmapatentsblog.com
“provides a Cmax in a human subject equivalent to the Cmax obtained when the first of three doses of a standard immediate release formulation having one third the amount of guaifenesin is dosed.”
“fully bioavailable in the subject’s stomach”
This is worthless intended use and functional crap that should be ignored when construing a composition claim.
For a court that has a whole doctrine of them, this represents a pretty epic failure to understand equivalents.
“Requiring a 90% confidence interval would inappropriately raise the bar for establishing infringement.”
The court did not explain (other than 90% seems like rather a high number) how it reconciled this statement with Perrigo’s argument that “the 80 to 125% range ‘means absolutely nothing in terms of establishing bioequivalence under FDA’s guidelines without the 90% confidence interval, as, among other things, it is the confidence interval itself that must fall within the 80-125% range.'” (emphasis in original)
Also, if the court is correct that “Adams must show that it is more likely than not that Perrigo’s ANDA product will have a Cmax within the 80-125% range” (emphasis mine), then what is the standard for showing anticipation by the same product? This reasoning makes it possible that products may at once infringe and not anticipate, which also means you could infringe based on a product you’ve been selling since long before the patent was filed.
Comments are closed.