Apotex Invalidates Two Patents Relating to Eli Lilly’s CIALIS® (tadalafil)

The Goodmans IP Team, including Andrew Brodkin, Ben Hackett, Jordan Scopa and Jaclyn Tilak, successfully represented Apotex in two patent infringement trials relating to Eli Lilly’s blockbuster ED medication, CIALIS® (tadalafil). Apotex has been marketing tadalafil products for the treatment of ED in Canada since July 2016 and Eli Lilly has been pursuing a patent infringement action against Apotex since September 2016.

The first trial, which was held in December 2019 (with closing oral arguments held in February 2020), related to Canadian Letters Patent No. 2,371,684 (the “684 Patent”), which claims particular doses (2 to 20 mg) of tadalafil for the treatment of ED. Additional defendants included Mylan, Teva, Pharmascience and Riva.

The trial judge concluded that the asserted claims of the 684 Patent were invalid on the bases of anticipation and obviousness and thus that there could be no infringement by Apotex (or any of the defendants).

The second trial, which was held in January 2020 (with closing oral arguments held in February 2020), related to Canadian Letters Patent No. 2,492,540 (the “540 Patent”), which claims processes related to the manufacture of tadalafil (i.e., the active pharmaceutical ingredient (“API”)). By the time of trial, the only remaining defendant in respect of the 540 Patent was Apotex.

The trial judge concluded that the asserted claims of the 540 Patent were invalid on the bases of anticipation and/or obviousness and thus that there could be no infringement by Apotex. In addition, even if the 540 Patent were valid, the trial judge determined that Eli Lilly failed to meet its burden to prove infringement in respect of three of the four processes utilized by Apotex’s suppliers to manufacture the tadalafil API. Apotex did not contest that, if the 540 Patent were valid, one of the four processes infringed some of the claims of the 540 Patent.

The 684 Patent Decision (Eli Lilly v. Apotex et al., 2020 FC 816)

The 684 Patent describes and claims doses of 2 to 20 mg of tadalafil for the treatment of ED.

In the result, the trial judge agreed with the prior conclusions of the Federal Court, in the context of an application pursuant to the former version of the Patented Medicines (Notice of Compliance) Regulations, that the 684 Patent was anticipated and also rendered obvious by Eli Lilly’s earlier 784 Patent, which claimed the use of tadalafil to treat ED. See Eli Lilly v. Mylan, 2015 FC 125.

Having determined that the 684 Patent was not a selection patent (which was the same conclusion previously arrived at by the Federal Court), the trial judge focused her anticipation and obviousness analyses on the subject matter of the claims of the 684 Patent.

To avoid anticipation, as a matter of claim construction, Eli Lilly urged the trial judge to “read in” a claim limitation to a “maximum daily dose” or “one dose per day”, which the trial judge rejected, including because, during patent prosecution, Eli Lilly had removed a maximum daily dose from the claims in response to an objection from the patent examiner.

Having construed the claims as not including a “maximum daily dose” or “one dose per day”, the trial judge concluded that the essential elements of each of the asserted claims of the 684 Patent had previously been disclosed and enabled by Eli Lilly’s 784 Patent. In particular, the 784 Patent disclosed that a pharmaceutical dosage form (such as a tablet) containing tadalafil in a range between 0.2-400 mg could be orally administered to treat ED. This necessarily meant that the 784 Patent also disclosed that a pharmaceutical dosage form (such as a tablet) containing tadalafil in a range between 2 to 20 mg could be orally administered to treat ED (as claimed in the 684 Patent).

For purposes of obviousness, the trial judge considered and applied now-settled jurisprudence of the Federal Court of Appeal indicating that the obviousness analysis ought to focus on any differences between the state of the art and the subject matter defined by the claims (i.e., the essential elements of the claims, as identified by claim construction).

The trial judge determined that the sole difference between the prior art and the subject matter defined by the claims of the 684 Patent was the lower and narrower dosage range (2 to 20 mg) given that the prior art had already disclosed that oral administration of tadalafil is effective at treating ED with a unit dose between 0.2 to 400 mg.

Relying on the expert evidence tendered by the defendants, the trial judge then concluded that it would have been a matter of “routine work” for the skilled person to arrive at the dosage range of 2 to 20 mg. In particular, the skilled person would have used the available information relating to tadalafil’s potency, selectivity and bioavailability (along with the same information and dosing information for VIAGRA® (sildenafil)) to predict the doses with the best balance between efficacy and safety and tolerability and then would have proceeded with the typical dose ranging studies conducted in Phase II clinical trials to successfully identify the minimum effective dose of tadalafil. This routine work would have inevitably resulted in the skilled person identifying the doses of 2 to 20 mg claimed in the 684 Patent.

Accordingly, the trial judge declared that all of the asserted claims of the 684 Patent were invalid and dismissed the infringement action as it related to the 684 Patent.

The 540 Patent Decision (Eli Lilly v. Apotex, 2020 FC 814)

The 540 Patent describes and claims a manufacturing process to synthesize tadalafil and focuses on the synthesis of the key intermediate compound, a cis-diastereomer. The synthesis of this key intermediate is achieved by carrying out a Pictet-Spengler Reaction (“PSR”) in which the desired cis-diastereomer is insoluble at reflux temperature or lower, and the undesired trans-diastereomer is soluble at reflux temperature or lower, resulting in concomitant crystallisation, and separation, of the desired cis-diastereomer product.

Following the Federal Court’s decisions in Eli Lilly v. Apotex, 2009 FC 991, aff’d 2010 FCA 240 and Merck v. Apotex, 2010 FC 1265, aff’d 2011 FCA 363, the trial judge declined Eli Lilly’s invitation to apply the statutory presumption of infringement pursuant to section 55.1 of the Patent Act, which relates to patents granted for processes for obtaining a “new product”.

The trial judge’s claims construction disposed of infringement with respect to two of Apotex’s suppliers’ processes to manufacture the tadalafil API. In all instances, the trial judge applied the well-settled principles of claims construction and preferred the construction advanced by Apotex. Again, Eli Lilly attempted to advance a construction inconsistent with its position during patent prosecution and the trial judge rejected Eli Lilly’s inconsistent construction.

If the claims of the 540 Patent were valid, the first process was found to infringe claims 1, 3 and 4, which was uncontested by Apotex, and not to infringe the balance of the asserted claims. The second process was found not to infringe any of the asserted claims.

The third process was found not to infringe on the basis that all of the tadalafil API manufactured using that process was used for regulatory testing such that it was never tableted and sold in Canada or for export. The trial judge re-affirmed the broad, clear and unequivocal regulatory exception to infringement provided by section 55.2 of the Patent Act.

Regarding the fourth process, given that Apotex led evidence that it went to significant and reasonable lengths to locate and produce additional information regarding that process, the trial judge declined Eli Lilly’s invitation to rely on the “common law presumption” of infringement articulated in Hoffmann-La Roche v. Apotex, 1983 CarswellOnt 871 (ON HC), aff’d 1984 CarswellOnt 1197 (ON CA). The fourth process was thus found not to infringe on the basis that Eli Lilly’s own expert confirmed that it was non-infringing and Eli Lilly proved the process by tendering the portion of Apotex’s abbreviated new drug submission that described the fourth process as a business record.

For purposes of anticipation of claims 1, 3 and 4 of the 540 Patent, Apotex relied on a particular example in Eli Lilly’s 594 Application, which also described processes leading to the manufacture of the tadalafil API. The key issue regarding anticipation was whether the “resulting suspension” observed in the relevant example in the 594 Application was the desired cis-diastereomer claimed in the 540 Patent. The trial judge accepted the opinion of Apotex’s expert, Dr. Anderson, “that it is most likely, highly likely, almost certainly the case…the suspended solid...at the 73-hour mark were indeed the desired cis-diastereomer” and rejected Lilly’s argument that Apotex was required to conduct an experiment to determine the nature of the suspension.

For purposes of obviousness, Apotex relied on Eli Lilly’s 594 Application as well as Eli Lilly’s 377 Patent, which also described a process leading to the manufacture of the tadalafil API. The trial judge accepted Apotex’s Dr. Anderson’s evidence that the skilled person would have arrived at the subject matter defined by the claims of the 540 Patent by using typical reaction conditions and performing a routine solvent screening of a finite number of reaction solvents suitable for use in commercial manufacturing, including the claimed isopropyl alcohol.

Accordingly, the trial judge declared that claims 1 and 3 to 4 of the 540 Patent were invalid for anticipation and all of the asserted claims of the 540 Patent were invalid for obviousness and dismissed the infringement action as it related to the 540 Patent.

Author: Jordan Scopa

The author would like to thank Andre Lapierre and Matt Buck of the Goodmans IP Team for their tireless scientific support throughout these proceedings.