Vifor Kept Its Process In Range 90 Percent of the Time. The Federal Circuit Said That Meant Zero Infringement.
By Prashant Kotian, PhD Researcher (Chemistry), ICT Mumbai
Ferric carboxymaltose, sold as Injectafer, is one of the most widely used treatments for iron deficiency anemia. Getting iron into a form the body can absorb without free iron causing toxicity means binding it to a carbohydrate scaffold, and the specific chemistry used to build that scaffold is where this case starts.
U.S. Patent 8,759,320, owned by the Italian company Biofer S.p.A., covers a process for making that iron-sugar complex. The process runs in four steps: activate a sugar, complex the activated sugar with iron, purify it, then stabilize it. Only the first step, which both sides in this case called "oxidation," was at issue on appeal. In that step, a large sugar molecule such as dextrin or dextran, carrying an aldehyde group (-CHO) at one end, gets oxidized to a carboxylic acid (-COOH) at that same end. The oxidizer is bromine, generated in situ from hypochlorite and a metal bromide rather than added directly, dosed "instant by instant" so no excess ever builds up in solution.
The underlying chemistry problem is selectivity. Dextrin and dextran are large molecules with more than one site that looks like an aldehyde to an oxidizer, not just the one at the chain's end. Oxidize the wrong site and the result is not a clean end-group conversion but "depolymerization," the sugar chain breaking apart. Biofer's own patent, backed by a declaration its named inventor Dr. Egidio Marchi submitted to the patent examiner, tied that outcome to one number range: running the reaction at pH 7.8 to 8.2 was highly selective for the terminal aldehyde, while pH 9.8 to 10.2 caused a depolymerizing attack on the chain instead. That comparison was not incidental. Biofer used exactly this data to persuade the examiner that the claimed pH range was critical enough to justify allowance.
Which is why claim 1's language, "reacting a sugar . . . in a solution at a pH between 7.0 and 9.0," became the entire dispute on appeal. Biofer sued Vifor, the Swiss company that markets Injectafer, arguing this same process claim covered Vifor's manufacturing. Vifor's actual production data, confirmed by Biofer's own testing during discovery, showed the reaction sat inside the 7.0 to 9.0 window for about 90 percent of the oxidation step, and drifted outside it for the remaining 10 percent.
Whether that 10 percent mattered turned entirely on how to read four words. Biofer's position at the district court was that "pH between 7.0 and 9.0" simply names two numbers bounding a range: if the reaction ever touches that window, even for an instant, the limitation is satisfied. The district court rejected that reading and construed the claim to require the pH be maintained inside 7.0 to 9.0 for the full duration of the oxidation step. It relied on three things: the sentence structure of the claim itself, where the pH clause modifies the entire reacting step rather than a moment within it; a specification passage stating the pH is "controlled and maintained" throughout the reaction; and, most decisively, Biofer's own prosecution history, where its inventor's declaration argued the claimed range was what prevented depolymerization. That argument only makes chemical sense if the range holds for the reaction's whole duration, not for an instant.
Under that construction, Vifor's 90 percent compliance figure was not a partial defense, it decided the case. If the claim requires pH to stay in range throughout the step, a process that drifts outside it 10 percent of the time simply is not the claimed process. Biofer stipulated to non-infringement rather than dispute the underlying data and carried the claim construction question up to the Federal Circuit instead.
On appeal, Biofer tried a fallback reading: the reaction runs mostly between pH 7.0 and 9.0 with brief excursions outside it, and that should still count as long as the reaction achieves the claimed "specific oxidation" by the end. The Federal Circuit never reached the merits of that theory. It held Biofer had never made this argument to the district court, where Biofer had in fact argued close to the opposite, that touching the range for "merely a millisecond" should be enough. A party cannot swap in a new claim construction theory on appeal that it never raised below, so the panel treated the argument as forfeited.
The court separately declined to review a second disputed claim term, over how "stoichiometric" the hypochlorite addition needs to be, because the parties' stipulated judgment never turned on that term to begin with. There was nothing live left to decide on it.
Nobody in this case argued about whether Biofer's chemistry and Vifor's chemistry were different once the claim construction question was settled; both sides agreed on the underlying facts immediately once the pH ruling came down. The entire two-year fight, and the entire appeal, rode on four words in one clause of one claim, and on prosecution-history statements Biofer's own inventor made a decade earlier to get the patent granted in the first place. For any process patent that recites a numerical reaction condition, whether pH, temperature, or concentration, the lesson is specific rather than general: whatever an inventor told the examiner about why a range matters is exactly what a future defendant will use to argue the range has to be maintained throughout, not merely visited.
What we do
See if this applies to one of your matters
We run genus/species (Markush) claim coverage analysis, Section 3(d) compliance screening, and prior-art structural triage on real chemical structures — not keyword search. If a case like this one is live for you, we'll run a free sample on a compound of your choosing.