Apotex' earliest potential launch date for its generic version of GSK's Paxil is September 2003.

Philadelphia federal court Judge Barclay Surrick granted Apotex' motion for summary judgment of noninfringement on product-by-process claims in two Paxil patents (6,113,944 and 6,172,233) Dec. 20. The court denied Apotex' petition on a patent for a different form of paroxetine (5,900,423), however, and dismissed all but one claim on another (6,080,759).

GlaxoSmithKline has received five 30-month stays since May 1998 on Paxil patents; the last stay (for the '233 patent) expires in September 2003, likely the earliest date FDA could approve Apotex' generic paroxetine (see chart: " ¹ GSK Stays Off Paxil Generics "). FDA granted tentative approval to Apotex' ANDA for the antidepressant in May 2001.

All four Paxil patents in the suit were late-listed by GSK. Apotex' first 30-month stay (triggered by the '723 patent) expired in November 2000, and the second (for the '423 patent), expired in January 2002. Stays for the '759 and '944 patents will expire in February 2003 and June 2003, respectively.

The paroxetine hydrochloride compound patent (4,077,196) expired prior to Paxil's December 1992 approval, but SmithKline Beecham (now GSK) listed another patent in the "Orange Book" (4,721,723), which claims paroxetine hydrochloride hemihydrate.

GSK's multiple 30-month stays for Paxil were questioned by the Federal Trade Commission in a July generic drug report. The late-listed patents belong to categories (including product-by-process) that "raise significant listability issues," FTC said (² Also see "Paxil Late-Listed Patents Are Questionable For Orange Book Listing – FTC" - Pink Sheet, 5 August, 2002., p. 8).

FDA is considering whether product-by-process patents require special declarations for listing in the Orange Book as part of a proposed rule.

In his decision for summary judgment, Judge Surrick found that GSK's product-by-process claims in the '944 and '233 patents are anticipated by the '723 patent and are therefore invalid.

In arguing against Apotex' motion, GSK maintained that the dry amixing process covered by the '944 patent results in tablets that are "identifiably different" than those made by other processes. Product made by wet granulation, for example, is more likely to develop a pink hue, retain spherical granules, and have different content uniformity, the company said.

The court disagreed, noting that the '944 patent specification does not mention the absence of spheres and content uniformity. While the "pink hue" is referenced in the patent, it does not state that the color is eliminated by using the process, the decision says.

"In short, the product characteristics now cited by SmithKline are insufficient to distinguish the product of the '944 patent from the products claimed in the '723 patent," the court determined.

In arguing that Apotex infringed the '233 product-by-process patent, GSK maintained that the tablets resulting from the process described contain a "distinctive 4-fluorobiphenyl component that is not present in the prior-art paroxetine products."

Judge Surrick again disagreed, stating that the court found no reference to the 4-fluorobiphenyl impurity in the '233 patent. "SmithKline may not use process claims to define its product based on distinguishing characteristics identified after the fact."

The court refused to grant Apotex summary judgment on the '423 patent, and all but one claim from the '759 patent. Both cover paroxetine hydrochloride anhydrous Form A: the '423 patent covers the form itself, while '759 refers to the process for making the compound.

Apotex argued that summary judgment is defensible for the two patents because the melting point of its proposed generic paroxetine is lower than that of GSK's product. Form A has a melting point of about 123° C to 125° C, while Apotex' ANDA specifies a melting point of 116° C to 120° C.

GSK countered that the melting point of Form A could be lower in certain circumstances, in which case Apotex would infringe its patent.

The '423 patent specifications state that the 123° C to 125° C melting point is achieved only with .05% bound solvent. Form A is described as typically having less than 2% bound solvent, therefore solvent with more than .05% but less than 2% could have a lower melting point, GSK argued.

Judge Surrick agreed with GSK: "After considering the parties' submissions, particularly regarding the significance of the Form A melting point as it relates to construction of the patents and the issue of literal infringement, we are not prepared to grant summary judgment at this time."

The court did agree with Apotex that one product-by-process claim in the '759 patent (claim 6) is invalid based on the principles of collateral estoppel. An earlier court decision found claims six through nine to be invalid on the grounds of double patenting, because they address the same invention described in claim one of the '423 patent.

Apotex views the decision as a victory, noting that only one Paxil product claim remains: the '423 claim for paroxetine hydrochloride anhydrous Form A.

Apotex "intends to accept Judge Surrick's invitation to renew its noninfringement summary judgment motion on the remaining product and process claims after Judge Surrick formally construes the meaning of the remaining product claims," the company said.

"We realize there still are a few legal hurdles to clear, but we are hopeful that the courts' willingness to strike down invalid patents will discourage some brand-name drug makers from manipulating FDA regulations by serially listing patents to block the marketing of a less costly version of their products."

GSK indicated that it views the judgment more as a split decision, and in a Dec. 30 press release said it "continues to believe there are significant hurdles that prevent launch of a generic Paxil product." The company plans to appeal the claims held invalid.

Apotex Paxil Generics Earliest Launch Date Could Be September 2003

More from Archive

More from Pink Sheet