Not So Smooth of a “Transition”: FDA Sued Over Deemed BLA Transitions – or Lack Thereof

We’re a little late to the party on this one (let’s just blame it on social distancing), but, after 10 years in the making, Transition Day has finally come and gone for protein products.  For the uninitiated, on March 23, 2020, all products approved as drugs under a New Drug Application that meet the definition of a “protein” officially became biologic products known as “deemed BLAs” in accordance with the Biologics Price Competition Act of 2009 (“BPCIA”).  Initially, these products were to exclude “chemically synthesized polypeptides,” but a last minute revision extended the transition to all protein products, defined as “any alpha amino acid polymer with a specific defined sequence that is greater than 40 amino acids in size.”  After a 10 year transition period, around 100 products officially transitioned and are now eligible to be reference products for biosimilars (with no exclusivity other than existing orphan drug exclusivity).  FDA’s electronic Orange Book database has now been purged of all “deemed BLAs” that transitioned on March 23, 2020.  However, for posterity, FDA has preserved the pre-transition Orange Book in its BPCIA edition, which provides all of the patent and exclusivity information that was current as of March  20, 2020.

But, as to be expected, not everyone is happy about the way FDA has handled this transition.  As soon as protein products “transitioned,” Teva was ready with a new lawsuit challenging FDA’s decision not to transition its Copaxone (glatiramer acetate) product to a deemed BLA.  According to its labeling, Copaxone, approved under NDA 020622 in 1996, is a product consisting of “acetate salts of synthetic polypeptides, containing four naturally occurring amino acids” that is indicated for the treatment of relapsing forms of multiple sclerosis (MS).  Two generic versions referencing Copaxone are on the market: one approved in 2015 and the other in 2017.

Back in February 2020, Teva submitted Comments on the Draft Guidance for Industry: The “Deemed to be a License” Provision of the BPCI Act – Questions and Answers” outlining its argument that Copaxone meets the statutory definition of “biological product.”  Teva’s Comments rested heavily on the chemistry of Copaxone’s active ingredient, glatiramer acetate.  Glatiramer acetate, Teva explained, is a “complex mixture of polypeptides having an overall standardized size and proportion of amino acids.”  While glatiramer acetate is chemically synthesized, when Congress repealed the BPCIA’s exclusion of chemically synthesized products from the definition of protein, Copaxone become eligible to be deemed a biological product.  And because the polypeptides have an average length of 40 to 100 amino acids, Teva argued, Copaxone is squarely within FDA’s definition of a biologic.  FDA did not respond directly to this comment.

As we approached the official “transition” date, FDA adopted its final list of “Approved NDAs for Biological Products That Were Deemed to be BLAs on March 23, 2020.”  Copaxone was not on it.  And while FDA said it would send letters to all sponsors of transitioning NDAs, Teva received no such letter for Copaxone.  Teva, therefore, filed suit against FDA.

Teva’s lawsuit closely tracks its Comments submitted to FDA in February 2020.  Like the Comments, the Complaint explains that glatiramer acetate is a chemically synthesized polypeptide with an average length of 40 to 100 amino acids.  Noting that FDA’s proposed definition of “protein” requires the amino acid polymer to have “a specific, defined sequence,” the Complaint suggests that FDA did not transition Copaxone because of a lack of a “specific, defined sequence.”  Indeed, as Teva acknowledges, FDA had previously stated in another context that “glatiramer acetate is distinguishable from proteins because ‘it does not … have a defined and specific sequence.’” Since making that statement, Teva notes, FDA has characterized products with similar properties as Copaxone as proteins, including Vitrase (hyaluronidase for injection) and Creon (pancrelipase).  Therefore, Teva concludes that Copaxone should be treated in the same way as similarly situated products – as a deemed BLA.  Further, even if FDA disagreed with this position, Teva argues that Copaxone should still have transitioned because, “at a minimum, it fits squarely into the catchall category of ‘an analogous product.’”

The Complaint alleges that FDA’s failure to transition the Copaxone NDA into a deemed BLA is arbitrary, capricious, and contrary to the BPCIA in violation of the Administrative Procedure Act.  Because FDA’s adopted definition of protein focuses only on the number of amino acids in the polymer, a definition that Copaxone inarguably meets, Teva contends that Copaxone must be transitioned to a deemed BLA.  In the alternative, Teva argues that Copaxone is an analogous product because it has a similar structure to myelin base protein—a transition product.  And because the plain language of the BPCIA states that that a previously approved NDA for a biological product “shall be deemed to be a license for the biological product under [section 351 of the PHSA] on the date that is 10 years after the date of [the BPCIA’s enactment],” FDA has no discretion not to transition a product meeting the definition of a biological product.  It therefore must convert a qualifying NDA for a “biological product” to a deemed license.

In an effort to show the harm caused by FDA’s failure to transition Copaxone to a deemed BLA, Teva is candid about its motivations: enforcement of its process patents.  As we explained several weeks ago, the biosimilar patent dance includes process patents in the exchange of patent information, permitting the enforcement of process patents prior to biosimilar launch.  This contrasts with the NDA/small molecule context, in which the act of submitting an ANDA is only an artificial act of infringement with respect to a drug substance or drug product patent—meaning that a process patent is not infringed until launch.  Teva’s Complaint explains:

Had FDA transitioned the COPAXONE NDA to a deemed license, Plaintiffs could assert the Teva Ltd. process patents against any biosimilar applicant based upon the filing of a biosimilar application, because the filing of the application is an artificial act of infringement with respect to (inter alia) the process patents.  Under the ANDA process, by contrast, Plaintiffs cannot sue for an artificial act of infringement based upon the filing of an ANDA, because Teva Ltd.’s patents are not for “a drug” or “the use of [a drug].” 35 U.S.C. § 271(e)(2)(A). Instead, Plaintiffs must wait to sue until a generic applicant actually infringes a patent by “us[ing]” or “sell[ing]” the patented process; or “sell[ing],” “offer[ing],” or “import[ing]” glatiramer acetate made with the patented process; or until such an infringement is certainly impending. Id. § 271(a), (g).

This distinction allows biosimilar manufacturers more leeway for pre-launch patent enforcement.  This is important because launch of a lower cost alternative is typically a bell that cannot be unrung: once a product is launched, the holder of an infringed patent can recover monetary damages, but it’s almost impossible to fully remove already-launched infringing product from the market.  Therefore, enforcement of a process patent prior to launch has tangible benefits for sponsors, particularly where the manufacturing process may be integral to the safety and efficacy of a product.

Unsurprisingly, generic manufacturers have intervened.  Sandoz filed an Unopposed Motion to Intervene on April 17 citing the implications a Copaxone transition would have on Sandoz’s  approved ANDA.  Sandoz brought its ANDA to market in 2015 after 10 years of patent infringement litigation with Teva (which somehow included litigating process patents).  As an ANDA, Sandoz could theoretically recoup that investment, as a generic can be automatically substituted for a therapeutically equivalent brand product, but biosimilars “do not enjoy the same market advantages as therapeutically equivalent ANDA products,” causing Sandoz to “lose its ability to compete effectively in the market with Copaxone.”  And, of course, it would deny MS patients automatic substitution of an affordable alternative.

Teva’s argument—that the plain language of a statute limits FDA’s discretion—is one that has had some recent success against FDA.  Genus Medical Technologies recently prevailed against FDA on the question of whether FDA has discretion to regulate a device as a drug.  Spoiler alert: it does not (though FDA has appealed).  This question here is similar.  However, there is a lot of room here for FDA to argue that Copaxone does not meet the definition of “biologic;” whereas, in the Genus matter, FDA expressly admitted that the product met the definition of a device.  As such, there is a possibility that a court could view this more of a question of science than discretion or fair application of a statute.  And courts often defer to FDA in matters of science.  Adding another wrench in this case is that it has been assigned to Chief Judge Beryl A. Howell, who recently held that FDA’s application of the three-year exclusivity statute was arbitrary and capricious, suggesting that FDA does not necessarily win on Chevron Step 2 in her court.  This could be a close case, and we’re looking forward to seeing how it plays out.