Can Amicus Therapeutics stop entry of Aurobindo’s generic Galafold?
Zachary Silbersher
Amicus Therapeutics has been fighting off prospective generics for Galafold®. Teva settled with Amicus in October 2024, but Aurobindo continued to fight. The parties recently went to trial, and on November 5, they exchanged their first round of post-trial briefing. The briefing suggests Amicus faces risk under Section 101 for at least two of its patents as well as prior art risk on all of them. What do the briefs say?
Amicus sells Galafold® (migalastat), which is a treatment for Fabry disease. Patients with Fabry disease often have a mutated GLA gene that leads to a defective or absent alpha-Gal A enzyme. Numerous different mutations in the galactosidase alpha gene (GLA) can lead to Fabry disease. Galafold is indicated for patients with Fabry disease who have a specific genetic variant in the GLA gene that is responsive to Galafold. Galafold binds to the defective enzyme and helps it travel into the lysosome. Galafold, however, only works with GLA mutations that are amenable or responsive to Galafold.
Amicus has asserted three patents against Aurobindo, including U.S. Patent Nos. 11,633,388 (“the ‘388 patent”); 12,042,490 (“the ‘490 patent”); and 11,833,164 (“the ‘164 patent”). Each patent is generally directed to a method of treating Fabry disease by administering migalastat. Two of the patents, the ‘388 and ‘490 patents, require that the patient suffering from Fabry disease also “has an α-galactosidase A protein comprising a HEK assay amenable mutation,” wherein the mutation is selected from a specifically enumerated group. The third patent, the ‘164 patent, similarly requires “identifying one or more α-galactosidase A mutations” from a specifically enumerated group before administering migalastat to the patient.
Aurobindo’s post-trial brief indicates that it is hanging its hat on invalidating Amicus’ patents. Its brief does not include any non-infringement arguments. Aurobindo makes two invalidity arguments. First, Aurobindo argues that Amicus’ patented methods of treating Fabry disease are invalid under Section 101 of the Patent Statute.
Section 101 proscribes ineligible types of subject matter for patents. For instance, patents cannot claim natural laws. Einstein could not have patented his admittedly genius discovery of the natural law that E=mc2. Otherwise, patents could be used to prohibit competitors from using natural laws, which in turn would presumably hinder rather than encourage scientific innovation.
Aurobindo argues that Amicus has essentially patented a natural law. Treating Fabry patients with migalastat was already in the prior art. Instead, Amicus only discovered that the drug works best with patients who have a certain pre-existing genetic mutations. According to Aurobindo, Amicus’ patents don’t do more than patenting that natural-law discovery.
Aurobindo previously moved for summary judgment that Amicus’ patents were invalid under Section 101. In September 2025, before the trial, the district court denied the motion, but did not prohibit Aurobindo from raising it again at trial. The court indicated the parties had not presented sufficient argument on whether Amicus’ patents were directed to an abstract idea—i.e., were the patents directed to an affirmative method of treatment? Or were they really directed to just a natural phenomenon—i.e., the idea that migalastat only works for patients having certain GLA mutations?
The court also admonished the parties for failing to address the Federal Circuit’s decision in, INO Therapeutics LLC v. Praxair Distribution Inc., 782 F. App’x 1001 (Fed. Cir. 2019). In that case, the patents were directed to treating infants with hypoxic respiratory failure with iNO gas, but included a proscription against administering the gas to patients with a congenital heart condition known as LVD. The inventors had discovered that administering iNO gas to patients with LVD could be life-threatening. The Federal Circuit held the patents were ineligible under Section 101. The prior art already taught administering iNO gas to infants with hypoxic respiratory failure. The Court found that adding a proscription against administering the drug to infants with LVD was just an instruction not to act, and such an “exclusion step merely restates the natural law.” (Slip Op. at 10).
Like the patents in INO Therapeutics, Aurobindo’s post-trial brief suggests that Amicus’ patents do nothing more than claim a natural law and should therefore be deemed ineligible under Section 101. Importantly, Aurobindo argues that “the asserted claims describe dosing methods that do not change depending upon the specific mutation or its amenability to migalastat.” For the ‘388 and ‘490 patents, the administration step is the first step. That administration step is not dependent upon first determining that the patient has the amenable GLA mutation.
Given that, Aurobindo has a tractable argument that these two patents are directed to ineligible subject matter. Ino other words, administering migalastat to Fabry patients was already in the prior art, yet the claim only adds a determination that the patient has a certain genetic mutation. If that determination does not affect whether or not the drug is administered, then the patent arguably claims nothing more than the natural law, just like in the INO Therapeutics case.
Another case from the Supreme Court, also cited by Aurobindo, is also arguably on point. In Mayo Collaborative Servs. v. Prometheus Labs., Inc., 566 U.S. 66 (2012), the patent claim did have an administering step, which recited, “administering a drug providing 6-thioguanine to a subject having said immune-mediated gastrointestinal disorder.” But the drug administration step in Mayo was the first step in the claim. It was not based upon prior testing. Rather, the drug was first administered to the patient, and next, the patient’s level of 6-thioguanine was measured, and the results of that measurement “indicates a need” to either increase or decrease the dosage. The Supreme Court held the patents invalid under Section 101.
On the other hand, with regard to Amicus’ third patent, the ‘164 patent, the drug is administered only after first determining that the patient has one of the requisite gene mutations. The final step of claim 1 recites, “administering migalastat or a salt thereof to the subject determined to have at least one mutation selected from the first group.” That suggests the patent does not simply claim the natural law, but bases the decision whether to administer a drug to a patient based on that natural law.
That difference between the patents may lead to invalidation of the ‘388 and ‘490 patents, but not the ‘164 patent. It will be interesting to see if Amicus argues this distinction in its opposition brief. Technically, Amicus needs only a single patent to block Aurobindo’s generic from the market. Given that the ‘164 patent does not expire until at least 2040, that may be all that Amicus needs.
Aurobindo also argues that Amicus’ patents are invalid against prior art. This is potentially where Amicus faces risk on all three patents. It appears that Amicus does not dispute that the prior art already taught using migalastat to treat Fabry disease at the claimed dosages. Rather, Amicus’ purported discovery was that the drug worked better for patients who have certain GLA mutations.
Aurobindo appears to have identified prior art that, at the very least, suggests that certain types of mutations are responsive to migalastat. The prior art doesn’t appear to identify each of the numerous mutations specifically recited in Amicus’ patents. On the other hand, the patented methods of treatment only require identifying a single mutation among those recited. If Aurobindo could pinpoint a single mutation identified in the prior art and recited in each of the patents, that could be enough to invalidate the patents under Section 103 of the Patent Statute.
Amicus argues only that its patents are non-obvious because of certain objective indicia of non-obviousness. In particular, Amicus claims that its patents satisfied a long-felt, but unmet need, which by itself should demonstrate the patents are not invalid. Amicus claims that before Galafold, the only available treatment for Fabry patients in the U.S. was ERT (enzyme replacement therapy). In other words, even though the prior art may have already directed scientists to explore specific mutations that responded to migalastat for treating Fabry disease, Galagold is purportedly a commercial embodiment of Amicus’ discovery of those particular mutations that satisfied an unmet medical need for Fabry patients.
Both parties will be submitting opposition briefing in a matter of weeks to respond to the other parties’ arguments. Most importantly will be whether Amicus can identify any specific limitations from the patents that are missing from Aurobindo’s cited prior art.