Date: 20101020
Docket: A-6-10
Citation: 2010 FCA 275
CORAM: EVANS
J.A.
DAWSON
J.A.
STRATAS
J.A.
BETWEEN:
BAYER SCHERING PHARMA
AKTIENGESELLSCHAFT
Appellant
and
THE ATTORNEY GENERAL OF CANADA
Respondent
REASONS FOR JUDGMENT
EVANS J.A.
A. INTRODUCTION
[1]
This is an
appeal by Bayer Schering Pharma Aktiengesellschaft (Bayer) pursuant to section
41 of the Patent Act, R.S. 1985, c. P-4 (Act), from a decision of the
Federal Court (2009 FC 1249), in which Justice Boivin dismissed its appeal from
a decision of the Commissioner of Patents, dated May 21, 2008. In that
decision, made pursuant to section 40 of the Act, the Commissioner refused to
grant a patent for a compound, because a patent had already been issued for the
same compound made by a particular process.
[2]
Bayer says
that the Commissioner erred in law in refusing to issue a patent. Counsel
argues that the jurisprudence establishes that a patent may be issued for a
product when a patent has been issued for the same product made by a particular
process. The general prohibition on issuing patents for inventions which, when
compared to an already patented invention, involve no ingenuity does not apply
to this situation.
[3]
In my
view, this cannot be correct because it is inconsistent with a fundamental
premise of Canadian patent legislation: a patent is a bargain in which an
inventor is granted a time-limited monopoly in the exploitation of an
invention, in exchange for fully disclosing it so that the public can use it
after the patent has expired. However, if a claimed “invention” is obvious, the
public derives no benefit from its disclosure and there is no justification for
granting the monopoly. If cases have decided that a patent may issue for a medicinal
compound when a product-by-process patent for the same compound has already been
issued, they are inconsistent with this principle and, in my opinion, should
not be followed.
B. FACTUAL BACKGROUND
[4]
The facts
of this case must be considered in light of provisions of the Patent Act concerning
medicines. At the relevant time, subsection 41(1) provided that a patent could
not be issued for “inventions relating to substances … produced by chemical
processes, and intended for … medicine”. However, the subsection also permitted
the grant of patents for medicinal compounds when made by a process that was
“particularly described and claimed” in the patent; these are known as
process-dependent patents or product-by-process patents. I attach as an
Appendix the full text of the statutory provisions referred to in these
reasons.
[5]
Subsection
41(1) was subsequently renumbered as subsection 39(1) and repealed by an
amendment to the Act (subsection 39 (1.1)) enacted on November 19, 1987, with
effect from four years later. Thus, regardless of when an application for a
patent was filed, a patent could be granted for a medicinal compound itself
after November 1991.
[6]
The facts
in this case are not in dispute. The patent Application under consideration has
a substantial history which, for present purposes, can be summarized briefly.
The Application was filed in May 1986, but claims divisional status from
another application (the parent application) which had been filed in August
1982. The parent application was amended so as to limit it to one invention in
accordance with section 36(1) of the Patent Act, and to remove claims
for compounds in order to comply with subsection 41(1). What remained in the
parent application was a process- dependent claim for a compound. The parent
application was allowed and a patent (the parent patent) was issued in February
1987.
[7]
When
filed in May 1986, the Application contained process-dependent claims for
compounds which were originally included in the parent application, but were
later removed from it for the reasons explained above. Since no patent had been
issued as of November 19, 1991, when the repeal of subsection 41(1) became
effective, Bayer was able to amend the Application so as to claim a medicinal compound,
which was materially identical to the compound claimed in process-dependent
form in the parent patent.
[8]
By virtue
of the transitional provision in section 78.1 of the Patent Act, the
Application is governed by the Act as it was immediately prior to October 1,
1989, the date when the Act was significantly overhauled.
C. THE DECISION OF THE
COMMISSIONER OF PATENTS
[9]
The
Commissioner agreed that Bayer was not applying for a second patent for the
same invention, because the parent patent, now expired, was, by necessity, a
product-by-process patent, while the claims in the new application were for the
medicinal compound itself. Accordingly, the Application was not refused on the
ground that it infringed the rule against the grant of more than one patent for
a single invention, which is often called “same invention” double patenting.
[10]
However,
the Commissioner rejected claims 1-12 in Bayer’s Application, on the ground of “obviousness”
double patenting. Each claim was dependent on claim 1. This was a claim for a
compound, which was the subject of process-dependent claims 1 and 8 of the
parent patent. Thus, claim 1 of the parent patent was for a process to produce
the compound, while claim 8 claimed the compound as produced by that process.
[11]
The
invention claimed in the Application (that is, the compound) was not “patentably
distinct” from the claim in the parent patent for the materially identical compound
when made by a particular process, and involved no “inventive ingenuity” or
novelty when compared to the claims of the process-dependent patent.
D. DECISION OF THE FEDERAL COURT
[12]
Bayer
appealed the Commissioner’s decision to the Federal Court pursuant to section 41
of the current version of the Patent Act.
[13]
The Court
held that the Commissioner’s decision was reviewable on a standard of
reasonableness, since the question in issue involved the interpretation of the Patent
Act, the statute administered by the Commissioner, and was within the scope
of her expertise. The Court held that reasonableness was also the standard of
review to the extent that determining whether a claimed invention is obvious is
a question of mixed fact and law.
[14]
The Court
then turned to the jurisprudence relied on by Bayer in support of the proposition
that the prohibition of “obviousness” double patenting does not apply to claims
for a product when an earlier patent has been issued for the same product made
by a claimed and described process.
[15]
It held
that the principal authority, Aventis Pharma Inc. v. Mayne Pharma (Canada) Inc., 2005 FC 1183, 42 C.P.R. (4th)
481, rev’d. on other grounds, 2008 FCA 21, 380 N.R. 35 (Mayne), dealt only
with “same invention” double patenting, not “obviousness” double patenting. Consequently,
the Court held, it was reasonable for the Commissioner to apply to the present
facts the general principle that a patent will not issue in respect of a claim
that, on the basis of inventions disclosed in existing patents, is obvious and
involves no inventive ingenuity.
E. ISSUES AND ANALYSIS
[16]
This
appeal raises two issues:
1. What is the standard of review
applicable to the Commissioner’s refusal to issue a patent for the compound
claimed in Bayer’s Applicaiotn?
2. Does the rule
prohibiting “obviousness” double patenting apply to an application for a patent
in respect of a product, when a process-dependent patent has already been
issued in respect of the same product?
Issue 1: Standard of review
[17]
Relying largely
on the presumption established in Dunsmuir v. New Brunswick,
2008 SCC 9, [2008] 1 S.C.R. 190 (Dunsmuir) and Canada
(Citizenship and Immigration) v. Khosa, 2009 SCC 12, [2009] 1 S.C.R. 339, that
a specialized tribunal’s interpretation of its enabling legislation is
reviewable on a standard of reasonableness, the Judge concluded that this
standard applied in the present case to the Commissioner’s interpretation of
the Patent Act.
[18]
With all
respect, I cannot agree. Dunsmuir teaches (at para. 62) that where
previous case law has satisfactorily determined the standard of review to be
applied to a tribunal’s interpretation of its enabling legislation, courts
should normally adopt that standard without indulging in a protracted standard
of review analysis.
[19]
Cases
decided both before and after Dunsmuir have held that the Commissioner’s
interpretation of the Patent Act is reviewable on a standard of
correctness: see, for example, CertainTeed Corp. v. Canada (Attorney
General), 2006 FC 436, 50 C.P.R. (4th) 177 at paras. 26-27; Aventis Pharma Inc.
v. Pharmascience Inc., 2006 FCA 229, [2007] 2 F.C.R. 103 at para. 20; Belzberg v. Canada (Commissioner of Patents), 2009 FC 657, 307 D.L.R. (4th)
664 at para. 34; Amazon.com. v. Canada (Attorney General), 2010 FC 1011 at paras. 28-29.
[20]
Two
considerations indicate that the case law has satisfactorily resolved the
standard of review applicable in this case to the Commissioner’s refusal to issue
a patent to Bayer for its chemical compound.
[21]
First, although
ultimately rooted in a fundamental principle of the Patent Act, the rule
against double patenting has in large part been made by judges on a case by
case basis, and is designed to prevent the “evergreening” of an invention by
unduly prolonging the period of the monopoly of its use through multiple patents
that contain only obvious or uninventive variations on the invention. Whether
the rule against double patenting is applicable here depends principally on the
interpretation of judicial decisions. The Commissioner has no more expertise
than the Court in deciding this question of law.
[22]
Second,
suppose that in the instant case the Commissioner had granted Bayer a patent
for the compound, and that Bayer subsequently brought an action for
infringement against a generic drug company producing it or, as was the case in
Mayne, sought a prohibition under section 6 of the Patented Medicines
(Notice of Compliance) Regulations, SOR/93 133 (PMNOC Regulations). The generic
could have defended the proceeding by alleging that the patent was invalid for “obviousness”
double patenting, because of the previous process-dependent patent for the same
compound. Bayer could have argued, as it does in the present case, that the
rule against double patenting does not apply in this situation.
[23]
While
there is a statutory presumption that, in the absence of evidence to the
contrary, a patent granted is valid (subsection 43(2) of the current Patent
Act), the judge in this hypothetical would have had to determine for
herself whether the patent for the compound was valid as a matter of law, without
asking whether the Commissioner’s decision to issue it was reasonable. On
appeal, this Court would review the judge’s decision on questions of law on the
correctness standard: Halford v. Seed Hawk Inc., 2006 FCA 275, 275
D.L.R. (4th) 556 at para. 39.
[24]
If a
correctness standard is applied to a question of law when it arises in either a
trial or a prohibition proceeding under the PMNOC Regulations, there is no
basis for applying a reasonableness standard when the same question arises on
an appeal from the Commissioner: compare SOCAN v. Canadian Assn. of Internet
Providers, 2004 SCC 45, [2004] 2 S.C.R. 427 at paras. 48-50. The applicable
standard of review cannot depend on whether a patent was granted or refused, or
on the nature of the proceeding in which the question of law arose.
[25]
For this
reason as well, the Commissioner can claim no more expertise than the courts in
determining the legal question at issue here: does the rule against “obviousness”
double patenting apply on the present facts.
[26]
The
question at issue in this appeal has no factual element because the parties
agree that the compound claimed in Bayer’s Application and the compound in the
process-dependant claims of the parent patent are materially identical: see
para. 42 of Bayer’s memorandum of fact and law. It follows from this that the
invention claimed in the Application involves no inventive ingenuity when
compared to the process-dependent claims made for the same compound in the
earlier patent. Hence, it cannot be argued that reasonableness is the
appropriate standard of review because the question decided by the Commissioner
was one of mixed fact and law.
[27]
Thus, in
order to succeed in this appeal, Bayer need only establish that the
Commissioner’s view of the law was wrong. Hence, it may invite the Court to
substitute its view of the legal question at issue for that of the
Commissioner.
Issue 2: Does the rule
prohibiting “obviousness” double patenting apply to Bayer’s
Application for a patent for the compound?
[28]
The
rationale of the rule against double patenting was clearly explained in Whirlpool
Corp. v. Camco Inc., 2000 SCC 67, [2000] 2 S.C.R. 1067 (Whirlpool)
at para. 37, where, writing for a unanimous Court, Justice Binnie said (at
paras. 37 and 63):
A patentee
who can “evergreen” a single invention through successive patents by the
expedient of obvious or uninventive additions prolongs its monopoly beyond what
the public has agreed to pay. …
…
The
prohibition against double patenting relates back to the “evergreen” problem
mentioned at the outset. The inventor is entitled to “a” patent for each
invention: Patent Act, s. 36(1). If a subsequent patent issues with
identical claims, there is an improper extension of the monopoly.
[29]
Justice
Binnie also noted (at para. 66) that the rule against double patenting is not
limited to “same invention” double patenting, where an applicant seeks a patent
for the same invention as that claimed in an existing patent.
There is,
however, a second branch of the prohibition which is sometimes called
“obviousness” double patenting. This is a more flexible and less literal test
that prohibits the issuance of a patent with claims that are not “patentably
distinct” from those of the earlier patent.
See also Eli Lilly Canada Inc. v. Novopharm Ltd.,
2010 FCA 197 at paras. 65 and 73.
[30]
This passage
would seem apposite in the present case, because there is nothing inventive or
“patentably distinct” in a claim for a product that is the subject of a previous
process-dependent patent. On the other hand, a process-dependent patent may be
granted, even though a patent has already been issued for the product itself, if
the process claimed and described for making the product exhibits inventive ingenuity.
[31]
Despite Whirlpool,
Bayer says that decided cases establish that a patent may be issued for a
compound, even though a product-by-process patent has already been issued in
respect of the materially identical compound. Counsel relies on Mayne; Pfizer
Canada Inc. v. Canada (Minister of Health), 2008 FCA 108, [2009]
1 F.C.R. 253 (Pfizer); and Apotex Inc. v. Sanofi-Synthelabo
Canada Inc., 2008 SCC 61, [2008] 3 S.C.R. 265 (Sanofi-Synthelabo).
Bayer submits that the law excludes the situation under consideration in the
present case from the general prohibition of “obviousness” double patenting.
(i) Mayne
[32]
The
decision of the Federal Court in Mayne is the principal authority on
which Bayer relies. Its facts are materially indistinguishable from those of
the present case. Aventis had been granted a product-by-process patent for a
medicinal compound (the ’343 patent) when subsection 41(1) was still in force.
It was later granted a patent (the ’682 patent) for the same compound after the
Act was amended to permit patents to be issued for a medicinal compound itself.
[33]
Aventis
sought a prohibition under section 6 of the PMNOC Regulations to restrain the
Minister of Health from issuing a Notice of Compliance to Mayne for an
infringing medicine until the expiry of its ’682 patent for the compound.
[34]
The
Federal Court Judge rejected (at para. 76) Mayne’s argument that the ’682
patent was invalid for double patenting.
Since it is
widely recognized that two patents can exist in parallel, one on the medicine per
se and one on the process used to produce that medicine, the argument on
double patenting cannot stand as patent ’343 never provided a monopoly on the
substance per se.
[35]
The
Attorney General says that Mayne is distinguishable, on the ground that
it only considered “same invention” double patenting. Counsel submits that this
can be inferred from both the paragraph quoted above and the absence of any
discussion in the reasons of whether the claims of the ’682 patent involved any
ingenuity when compared to those of the process-dependent patent for the same
compound. Further, counsel argues, Mayne did not specifically allege “obviousness”
double patenting. In the alternative, counsel says that Mayne was
wrongly decided because it is inconsistent with the bargain theory of patent
law and the explanation of “obviousness” double patenting in Whirlpool.
[36]
On the
other hand, Bayer points out that Mayne’s Notice of Allegation was not
restricted to “same invention” double patenting, but alleged simply that the
patent for the compound itself was invalid for double patenting. Moreover,
since Mayne was decided after, and explicitly refers to, the Supreme
Court of Canada’s decision in Whirlpool, it is inconceivable that the
Court was not aware of the “obviousness” branch of the rule against double
patenting, even though Justice Beaudry did not mention it expressly in his
reasons.
[37]
On the
basis of the arguments advanced by the Attorney General, I am of the view that the
Court in Mayne did not intend to decide that, as a matter of law, the
rule against “obviousness” double patenting excludes a claim for a compound
when a process-dependent patent with respect to the same compound has already
been issued. However, if Mayne did so decide, it is, in my respectful view,
wrong. Bayer provided no principled basis for excluding the facts of both Mayne
and the present case from the “obviousness” branch of the rule against double
patenting.
(ii) Pfizer
[38]
Counsel
for Bayer also relies on statements by Nadon J.A. as authority for the
proposition that this Court has already decided that “obviousness” double
patenting does not apply to the present fact situation. I do not agree, for two
reasons.
[39]
First, the
patent for a compound in issue in Pfizer (the ’546 patent) claimed a selection
from compounds covered by the ’768 patent. Justice Nadon rejected (at para. 69)
the allegation that the ’546 patent was invalid for anticipation, obviousness
and double patenting because the generic had not challenged the validity of the
selection. Consequently, he found it unnecessary to examine the generic’s
allegations under the headings of anticipation, obviousness and double
patenting.
[40]
Nonetheless,
Justice Nadon decided (at para. 69) to “say a few words regarding the issues of
double patenting and anticipation.” Since he explicitly recognized that his
“few words” on double patenting were not necessary for the disposition of the
appeal, they are obiter dicta and not binding as a matter of precedent.
[41]
Second, he
examined the generic’s argument that the ’546 patent was invalid for double
patenting because another patent, the ’441 patent, claimed processes for making
the compound claimed in the ’546 patent. Justice Nadon said (at para. 76):
In my opinion,
the double patenting allegations are not justified. The ’441 patent covers
processes whereas the ’546 patent covers compounds. As explained by Hughes
& Woodley [on Patents, 2nd ed. loose leaf (Markham,
Ont.: 2005, LexisNexis Canada)], (at §15, page 172), “[a] previous patent for a
product by a claimed process does not invalidate a later patent for the product
alone for reasons of double patenting” (see: Aventis Pharma Inc. v. Mayne
Pharma (Canada) Inc., …at paras. 72-76).
[42]
In my
opinion, the comparison in this passage of claims for a process in one patent and
claims for a compound in another suggests that Justice Nadon had “same
invention” double patenting in mind. He does not consider whether the ’546
patent was obvious when compared to the ’441 patent.
[43]
This view of
Justice Nadon’s reasons is confirmed by his agreement (at para. 78) with the
following statement of the relevant law in Pfizer Canada Inc. v. Canada
(Minister of Health), 2006 FC 1471, (2007), 54 C.P.R. (4th) 279. After
rejecting the argument that the ’546 patent for a compound was invalid for
“same invention” double patenting because the ’768 patent was in
process-dependent form, the Judge in that case, Justice von Finkenstein, said (at
paras. 101-102):
As far as “obviousness”
double patenting is concerned, the claims or
disclosure must exhibit
novelty or ingenuity in order for the second
patent to be valid ….
As the Court found that
the [second] patent was a selection patent, by
definition it is novel
and unexpected. It thus cannot be invalid on the
basis of “obviousness”
double patenting.
[44]
Thus, in
my opinion, it cannot be said on the basis of Pfizer that this Court has
decided that a patent for a compound is not invalid for “obviousness” double
patenting on the ground that a patent in process-dependent form has been
granted with respect to the same compound.
(iii) Sanofi-Synthelabo
[45]
Finally,
Bayer invokes the decision of the Supreme Court of Canada in Sanofi-Synthelabo,
where the Court held that a selection patent must exhibit inventive ingenuity.
The Court’s discussion of the prohibition of double patenting (at paras.
94-110) was principally directed to the “same invention” branch of the rule. However,
the Court also briefly commented on “obviousness” double patenting in the
context of selection patents. Thus, delivering the judgment of the Court,
Justice Rothstein said (at para. 113):
A selection patent that
claims a compound that is patentably distinct
from the genus patent
will not be invalid for obviousness double patenting.
[46]
Again, I
do not see how this assists Bayer in the present case, where it cannot be said
that the claims for the compound in the Application are patentably distinct or
exhibit any ingenuity when compared to those in the process-dependent patent
for the same compound.
[47]
Having
found that the Commissioner was correct to reject the Application for
“obviousness” double patenting, I need not address the other arguments advanced
by Bayer to support its appeal.
F. CONCLUSIONS
[48]
A claim
for a compound that is materially identical to the subject of a
process-dependent claim exhibits no inventive ingenuity or novelty. Since the
claim for the compound in Bayer’s Application is not materially different from
that in the process-dependent parent patent, the Application lacks novelty and
inventive ingenuity.
[49]
Consequently,
the Commissioner correctly denied the Application on the ground of “obviousness”
double patenting because the patent would have disclosed no new invention.
For her to have granted a patent for the compound would have
improperly “evergreened” the parent patent, now expired, by creating a second
monopoly in the use of the compound running from the date when the Application
was approved.
[50]
If Mayne
decided otherwise, it is, in my respectful view, incorrect and should not be
regarded as authority for the proposition that the prohibition of “obviousness”
double patenting excludes claims for a compound that has also been the subject
of a process-dependent patent.
[51]
For all these
reasons, I would dismiss Bayer’s appeal with costs.
“John
M. Evans”
“I
agree
Eleanor
R. Dawson J.A.”
“I
agree
David
Stratas J.A.”
Email this Content