The Federal Court released its reasons in Amgen Canada Inc v
Mylan Pharmaceuticals ULC on November 3, 2015. The
Court dismissed an application commenced under the Patented
Medicines (Notice of Compliance) Regulations (NOC proceeding)
by pharmaceutical company Amgen for an order prohibiting the
Minister of Health from issuing a market authorization (notice of
compliance) to Mylan for its version of Amgen's calcimimetic
drug SENSIPAR®. These
reasons are a rare instance of the Federal Court invalidating a
patent covering a chemical compound on prior art grounds. The
Court's judgment also serves as a reminder that verifying the
predicted properties of previously disclosed compounds is
undeserving of a patent.
Canadian Patent No. 2,202,879 (the 879 Patent) disclosed and
claimed a very large number of compounds, including the cinacalcet
molecule and its pharmaceutically acceptable salts.
Claim 5 was directed to cinacalcet, the medicinal ingredient
Mylan alleged that the 879 Patent was invalid for anticipation,
obviousness and obviousness-type double patenting. Amgen
acknowledged in the NOC proceeding that a prior art patent (the 828
Patent) disclosed cinacalcet generically as part of a claimed genus
of over a trillion structures, but Amgen argued that because the
inventors of the 828 Patent had neither made nor tested cinacalcet,
claim 5 of the 879 Patent was the "selection invention"
of cinacalcet from the compounds disclosed in the 828 Patent.
This selection, Amgen argued, was the "complete answer"
to Mylan's invalidity allegations.
Cinacalcet Not a Selection Invention
The Federal Court disagreed that cinacalcet and the 879 Patent
as a whole were selection inventions. Justice Phelan
reiterated that a selection patent does not differ from any other
type of patent, but he held that neither the 879 Patent nor
evidence submitted by Amgen supported that the invention of the 879
Patent was a selection: cinacalcet simply did not differ
qualitatively or quantitatively from other compounds disclosed in
the 828 Patent. In so doing, the Federal Court reaffirmed the
principle that it is not an invention to make and test compounds
that were previously disclosed, even generically or as part of an
enormous class, if the selected species do not have unique and
Having found that cinacalcet did not have novel properties, the
Federal Court held that the prior art 828 Patent anticipated claim
5 of the 879 Patent. Relying extensively upon Mylan's
experts, the Court concluded that there was no inventive difference
between the 828 Patent and cinacalcet. Mylan's experts
had opined that a skilled person would have made cinacalcet
"as part of a routine development based solely on the teaching
of the 828 Patent." Specifically, the skilled person had
to synthesize and test at most 200 compounds to arrive at
cinacalcet, and this testing was mechanical. On these
grounds, the Court concluded that cinacalcet was not novel over the
class of compounds disclosed in the prior art.
On the issues of obviousness and obviousness-type double
patenting, the Federal Court held that cinacalcet was not inventive
and that claim 5 was not patentably distinct from the invention
previously disclosed in the 828 Patent. Relying again on
Mylan's experts, the Court concluded that it would have been
obvious to try to obtain cinacalcet because a skilled person would
only need to make and test 200 compounds which, on the evidence,
was easy and mechanical. The simple verification that
cinacalcet had the properties described or predicted in the prior
art did not make it inventive.
Practically speaking, Amgen Canada v Mylan
PharmaceuticalsULC highlights the crucial role that
expert evidence plays in the Court's assessment of prior art
disclosures. The Court's reasons also confirm that the
generic disclosure of compounds as part of a large prior art class
will anticipate the alleged invention of a compound within that
class, unless the claimed compound has unique and unexpected
properties. Moreover, the Federal Court's conclusion on
the evidence that making and testing 200 compounds was simple and
mechanical illustrates that an alleged invention may be obvious to
try, and ultimately uninventive, if it only involves verifying
predicted properties through well-known and easy tests.
A recent Saskatchewan Court of Queen's Bench decision allowed a court-appointed receiver to sell and transfer intellectual property rights free and clear of encumbrances, finding that a license to use improvements of an invention was a contractual interest and not a property interest.
Register for Access and our Free Biweekly Alert for
This service is completely free. Access 250,000 archived articles from 100+ countries and get a personalised email twice a week covering developments (and yes, our lawyers like to think you’ve read our Disclaimer).