III. Background
A. The '380
Patent and Statins
[18]
Lovastatin, as made by the process of the '380 Patent, is
an example of a medicinally‑valuable drug that is produced by a process
of fermentation. In very simple terms, the laboratory begins with a
micro-organism - in this case, Aspergillus terreus – and, through increasingly
larger fermentations carried out in very controlled settings, manufactures the
API of interest.
[19]
The '380 Patent relates to “hypocholesteremic products from
the cultivation of a microfungus of the species Aspergillus.” Dr.
Antonio Gotto provided very helpful background information on the role of
“hypocholesteremic” medications, such as lovastatin, in the treatment of
cardiovascular disease. In addition to being qualified because of his stature
as a professor of medicine, Dr. Gotto’s experience as a treating physician
during the 1970s and 1980s was directly relevant to the matters before me.
[20]
Atherosclerosis is a type of cardiovascular disease that
occurs when cholesterol and other substances build up in the walls of arterial
blood vessels to form plaque. Over time, the build-up of plaque thickens and
hardens the arterial walls restricting the flow of blood from the heart. Heart
attacks and strokes may follow.
[21]
The build-up of plaque is promoted by low density
lipoproteins (referred to as LDL or “bad” cholesterol). According to Dr. Gotto,
the relationship between reducing “bad” cholesterol and reducing the risk of
cardiovascular disease has been known for over 20 years. Thus, a primary goal
of medicine is to lower LDL cholesterol. The class of drugs known as “statins”
are of great assistance in achieving this goal.
[22]
In Dr. Gotto’s words (Gotto Expert Report, Exhibit 2,
paras. 24, 38):
It was only with the
discovery of statins – starting with lovastatin (MEVACOR®) in the late 1970s –
that treatment of elevated cholesterol became much more effective.
. . .
The single most
significant discovery to date for the treatment of cholesterol was the
discovery of lovastatin in the late 1970s.
[23]
Dr. Gotto described how statins work to lower cholesterol.
Statins reduce the production of cholesterol by the liver. Specifically,
statins block the liver enzyme known as HMG-CoA reductase
(hydroxyl-methylglutaryl-coenzyme A reductase); hence, statins are known as HMG‑CoA
reductase inhibitors. Dr. Gotto made the general comment that “statins changed
medical practice” (Gotto Expert Report, Exhibit 2, para. 50). No one disagreed
with this opinion.
[24]
Lovastatin, as manufactured and sold by Merck (or its
predecessors in interest) under the trade name MEVACOR, was the first commercially-available
statin.
B. History of
AFI/Blue Treasure Production
[25]
An important part of the story for this litigation is how Apotex Inc.
became interested in lovastatin and how Apotex Inc., AFI and the Blue Treasure
Joint Venture became involved.
[26]
Dr. Bernard Sherman is currently the Chairman and Chief Executive Officer
of Apotex Inc., a company that he founded in 1973. During his oral testimony, Dr.
Sherman described Apotex Inc. in the following terms:
It's a pharmaceutical
manufacturer, the largest in Canada today. We produce primarily generic pharmaceutical
products, but also some innovative products. We have huge dosage form
manufacturing facilities. We are vertically integrated. We have chemical
plants. We spend enormously on research and development, the largest in Canada, and we
have divisions in many countries around the world and factories in many
countries, including chemical plants.
[27]
Apotex Inc. recognized the significance of the lovastatin market. Dr.
Sherman described lovastatin, in 1993, as “one of the biggest selling drugs in
the country at the time, close to $100 million a year”.
[28]
Of particular relevance to this litigation, Dr. Sherman told the Court how
his company’s version of lovastatin became entangled with a suddenly-changed
regulatory regime in 1993. Until 1993, it was possible for a generic company to
obtain a compulsory licence to allow it to produce a generic equivalent of a
patented medicine. The original intent of Apotex Inc. was to obtain a
compulsory licence to use Aspergillus terreus to make lovastatin.
According to Dr. Sherman, in 1993, the licence regime and licences issued
under it were cancelled. They were replaced with the PMNOC Regulations outlined
by Dr. Sherman as follows:
In 1993, not only were
the licences – the licence regime eliminated, including retroactive
cancellation of some licences – one applicable to this case – but, in addition
to that, a new regime was instituted, called the Patented Medicines (Notice of
Compliance) Regulations, pursuant to which patentees or first persons, persons
who had approval for the original brand, could list patents which they
purported were relevant to a product; and, if they listed the patent, then a
generic applicant, a second person, cannot get federal approval until the
requirements of those regulations are satisfied, which means that the second
person has to serve a notice of allegation in which it is alleged that the
patent will not be infringed or is invalid. Then, within 45 days, if the
patentee or first person institutes a prohibition application, which almost
always happens, there is a delay in federal approval until that matter is
resolved, which can take a very long time.
[29]
A relationship of interest to this case is that of Apotex Inc. and AFI. In
the mid-1980s, Apotex Inc. contracted with ABI Biotechnology Inc. (ABI) in Winnipeg to develop
and manufacture certain fermented products. Ultimately, the assets of ABI were
bought by Apotex Inc. and the company was renamed as AFI. Through AFI, Apotex
Inc. gained the capacity to manufacture products using fermentation processes. AFI
added to the vertical integration of the Apotex family of business entities.
[30]
AFI was to be the source of the API lovastatin. As described in detail by
Dr. Lasure, in her Expert Report (Exhibit 48), and by Dr. David Cox,
during his testimony, the following steps were taken by AFI:
·
AFI acquired Merck's deposited strains of Aspergillus terreus from
the American Tissue Culture Collection (ATCC), including a strain designated
ATCC 20542.
·
ATCC 20542 was then mutated by UV mutagenesis twice to create a mutant
strain of ATCC 20542.
·
AFI designated the strain as BN-2-70 and the process for manufacturing
lovastatin using this strain as AFI-1.
·
Between 1991 and 1995, AFI developed a commercial scale fermentation
process for making lovastatin using AFI-1 – that is, using Aspergillus
terreus.
[31]
In 1992, Dr. Sherman testified that, anticipating the intent of the
government, Apotex began to look for a non-infringing process. Apotex “had to
find a microbe that would produce lovastatin that was not Aspergillus
terreus”. In her Expert Report, Dr. Lasure summarized the context and the results
of this search:
·
On June 25, 1988, the Journal of Antibiotics published an article entitled
"The Synthesis of Compactin (ML-236B) and Monacolin K in fungi"
written by Dr. Akira Endo et al. Dr. Endo reported on fungal strains
capable of producing Monacolin K (known now to be lovastatin), including Phoma
species M4452.
·
In June 1992, a sample of Phoma Sp. M4452 was sent to AFI by Dr.
Endo.
·
For the next six months or so, AFI took steps in its laboratories to
confirm and develop the production of lovastatin from the Endo sample. The
process was initially referred to as Phoma #4; later designated as
“AFI-4”.
·
By May 1993, a sample of the AFI-4 product was confirmed to be Coniothyrium
fuckelii (also referred to as C. fuckelli).
·
Apotex Inc. filed a patent for the AFI-4 process – a process for making
lovastatin using Coniothyrium fuckelii – that subsequently issued as
United States Patent No. 5,409,820 on April 25, 1995.
[32]
As described by a number of witnesses, including Dr. Cox and Ms. Lori
Christofalos, AFI’s production of AFI-4 lovastatin and shipments to Apotex Inc.
can be divided into three phases:
1.
Phase 1 occurred between June 1996 and August 1997, during which all
production was done solely at AFI facilities in Winnipeg. The
finished API was shipped to Apotex Inc., beginning with the shipment of batch
CR0157 on December 2, 1996.
2.
In Phase 2, Blue Treasure (discussed below) manufactured approximately 70
batches of technical-grade lovastatin. The product was then shipped to AFI for
processing into API and shipment to Apotex Inc. Phase 2 lasted from about mid‑1997
to January 1998.
3.
Phase 3 consisted of approximately 294 batches of API-grade lovastatin
manufactured entirely at Blue Treasure after March 1998.The product was sent to
AFI, where “some testing” was carried out, and then shipped to Apotex Inc. This
phase continued until October 1999, with the last shipment received at AFI on
March 2, 2000.
[33]
The joint venture company known as Qingyuan Blue Treasure Pharmaceuticals
Co. Ltd. (Blue Treasure or Blue Treasure Joint Venture) is a critical component
of this litigation. Dr. David Cox, President and Chief Executive Officer
of AFI from September 1994 to September 1997, provided a clear and helpful
background about this joint venture. Dr. Cox was on the Board of Directors of Blue
Treasure during the same period.
[34]
Blue Treasure was formed pursuant to a Joint Venture Contract dated
January 25, 1994 among Qingyuan New North River Pharmaceutical Co. Ltd. (New
North River), Zuhai Special Economic Zone Lizhu Pharmaceutical Group Co. Ltd.,
Sichuan Industrial Institute of Antibiotics, AFI and BIOTECS. AFI held a 42.5%
share of the Blue Treasure Joint Venture. As set out in clause 4.01 of the
Joint Venture Contract, the purpose of the Blue Treasure Joint Venture was as
follows:
The purpose of the
Joint Venture is to renovate and operate the Factory, to purchase or otherwise
obtain all necessary raw materials and equipment required for the production of
the Products, and to produce, market, distribute and sell Products at a profit
to customers both in China and abroad.
[35]
In 1994, New North River was already an operational pharmaceutical
facility with capacity for carrying out fermentation processes. Under the Joint
Venture Contract, New North River contributed a portion of its facilities
located on its property to the Blue Treasure Joint Venture.
[36]
As defined in the Joint Venture Contract, “Products” meant “the drug
Lovastatin as Bulk Products and Finished Products”. Dr. Cox stated that the
Blue Treasure Joint Venture “was set up to produce and distribute and sell
Lovastatin in the Chinese domestic market”. AFI’s main contribution to the Blue
Treasure Joint Venture was the organism that produced lovastatin. Dr. Sherman
told the Court that the decision to move the production of lovastatin to Blue
Treasure was made for the following reasons:
[Blue Treasure] had
capacity there, and we wanted to move the production for Canada out of
Winnipeg, both to bring costs down and to free up Winnipeg to go on
for other things that would be needed later.
[37]
In the spring of 1995, AFI transferred to Blue Treasure the information
and knowledge it had developed to manufacture lovastatin made from Aspergillus
terreus. Among the things transferred to Blue Treasure were: a document
setting out the process for producing lovastatin from Aspergillus terreus,
entitled "Scale-up Process to 15000L Fermenter"; and, 25 vials of
strain BN-2-70 from seed bank A18-378, and 5 rice cultures from batch #CF0057
(shipped to Blue Treasure on May 15, 1995). Blue Treasure began producing
lovastatin, using the AFI-1 process, in 1996.
[38]
In about April 1997, AFI determined that it would transfer the AFI-4
technology to Blue Treasure together with a guarantee that AFI would purchase
the lovastatin from Blue Treasure, provided that the lovastatin was all made by
the AFI-4 process and that “the Blue Treasure facility be exclusively dedicated
to AFI-4”. The terms of this arrangement were set out in a letter agreement
dated April 16, 1997 between AFI and Blue Treasure. Dr. Cox described the
impact of the AFI-4 transfer as “transformative in a positive way”. The
transfer of AFI-4 to Blue Treasure was made with very explicit instructions
that the lovastatin purchased by Apotex was to be produced exclusively with the
AFI-4 Coniothyrium fuckelii strain, with no possibility of contamination
from Aspergillus terreus (see, for example, letter dated September 12,
1997 from Mr. Fowler to Mr. Zhou). Problems quickly arose. These problems are
discussed later in Section VII of these reasons.
[39]
From 1997 to 1999, AFI imported lovastatin from Blue Treasure in
accordance with the terms of the Blue Treasure Joint Venture. The lovastatin
API was then sold to Apotex Inc.
IV. Standing
[40]
The first issued raised by Apotex is the standing of Merck & Co. to
bring this action.
[41]
The authority of a party to claim damages for patent infringement is found
in s. 55(1) of the Patent Act.
|
55.(1) Any person who infringes a patent is liable to the
patentee and to all persons claiming under him for all damages sustained by
the patentee or by any person, by reason of the infringement.
|
55.(1) Quiconque viole un brevet responsable, envers le
breveté et envers toute personne se réclamant du breveté, des tous
dommages-intérêts que cette violation a fait subir au breveté ou à cette
autre personne.
|
[42]
The term “patentee” is defined in s. 2 of the Patent Act to mean
“the person for the time being entitled to the benefit of a patent”.
[43]
The '380 Patent was granted to Merck & Co. In 1985, Merck & Co.
entered into a License Agreement with Merck Frosst (the 1985 License Agreement),
granting an non-exclusive licence to Merck Frosst. That Agreement was amended,
effective January 1, 1989, to add the '380 Patent. Subsequently, as of January
1, 1992, Merck & Co. entered into an agreement (the MACI Agreement) with
Merck and Company, Incorporated (MACI) pursuant to which Merck & Co., as
Licensor, granted to MACI, as Licensee:
A permanent and
exclusive royalty-free license for the Intellectual Property which Licensor
owns or hereinafter acquires, but for any outstanding licenses for the
Intellectual Property which already granted pursuant to the License Agreement,
dated January 1, 1985, and amendments thereto between Merck & Co., Inc. and
Merck Frosst Canada Inc.
[44]
Apotex does not dispute Merck Frosst Canada Ltd.’s standing in this action,
as the successor in interest to Merck Frosst Canada Inc. However, Apotex
submits that Merck & Co. has no standing to bring this action, having
assigned all of its interest in the '380 Patent to MACI pursuant to the MACI
Agreement. Apotex asserts that, as of November 1992, MACI had the “full and
unrestricted benefit of the ‘380 Patent”. Merck & Co. lost all benefit of
the patent and, as a result, the right to damages under s. 55 (1) of the Patent
Act. Apotex argues that, although the agreement is entitled “License
Agreement”, a review of the words of the agreement demonstrates that the intent
of the parties to the MACI Agreement was to convey the entire right, title and
interest in the '380 Patent to MACI.
[45]
Apotex submits that agreements which take the form of a licence, but
nevertheless convey all of the substantive rights in a patent, have
consistently been held to constitute an effective assignment or transfer of
that patent. In support of this argument, Apotex relies on a line of
jurisprudence of courts in the United States and the United Kingdom (Merck
& Co., Inc. v. Francis R. Smith, 261 F.2d 162 at 164 (3rd
Cir. 1958); Vaupel Textilmaschinen KG v. Meccanica Euro Italia SPA, 944
F.2d 870, 874 (Fed. Cir. 1991); Prima Tek II, L.L.C. v. A-Roo Co., 222
F.3d 1372, 1377-78 (Fed. Cir. 2000); Guyot v. Thompson [1894] R.P.C. 541
at 554 (C.A.)[Guyot]).
[46]
I do not find the authorities relied on by Apotex to be of any assistance.
Except in the case of Guyot, above, a decision of the High Court of
Justice – Chancery Division, the Courts in those cases were considering the
effect of agreements in the context of U.S. patent
law. I do not see how they could guide this Court in determining the meaning of
the terms of and, if necessary, the intent of the parties to the MACI
Agreement. I did not have the benefit of an expert in U.S. law opining as to
whether the MACI Agreement would constitute a transfer of all of the rights of
the patent to MACI under applicable U.S. law. Moreover, the
facts in Guyot, where an exclusive assignee was attempting to enforce
the terms of an Indenture, are simply too remote from the question before me.
[47]
Rather, I would look at this issue in the context of the Canadian law of
contracts. As I understand the state of the Canadian law of contracts, the
express language of the parties to a contract is the core of their contractual
obligations. Where the words of a contract are clear and unambiguous, a court
need not look beyond those clear words to determine its intent and effect.
[48]
Apotex was unable to point me to a single Canadian case that supports its
position. Nevertheless, I would agree that the title of the License Agreement
would not be determinative if there is clear and persuasive evidence
that Merck & Co. intended to convey all of its rights in the '380 Patent
to MACI, retaining nothing to itself. Whether this is so or not will depend on
an examination of the words of the MACI Agreement and the facts and
circumstances surrounding the MACI Agreement.
[49]
In this case, the express language of clause 2 of the MACI Agreement uses
the word “license”. On its face, the MACI Agreement only grants a “license”.
The Supreme Court of Canada in Domco Industries Ltd., v. Armstrong Cork
Canada Ltd., [1982] 1 S.C.R. 907 at p.912, 66 C.P.R. (2d) 46, adopted the
comments of Fry L.J. at p. 470, in Heap v. Hartley (1889), 42 Ch. D.
461:
An exclusive license
is only a license in one sense; that is to say, the true nature of an exclusive
license is this. It is leave to do a thing, and a contract not to give leave to
anybody else to do the same thing. But it confers like any other license, no
interest or property in the thing. [Emphasis added.]
[50]
I also note the language of certain clauses in the MACI Agreement that
refer to rights retained by Merck & Co. For example, clause 3 provides the
Licensor with the rights to inspect the Licensee’s facilities. Under clause
5.2, the Licensee is to supply the Licensor with a detailed description of any
disclosure of “licensed know-how” to any governmental authority. In my view, retention
of rights such as these is inconsistent with an intention to transfer all
rights under the patent.
[51]
In support of its position, Apotex points to a recital to the MACI
Agreement:
WHEREAS, the Licensor
desires to grant the Licensee a permanent and exclusive license with respect to
its remaining right, title and interest in and to the rights which it
has acquired with respect to such intellectual property as a contribution to
the capital of the Licensee. [Emphasis added.]
Apotex relies on the Supreme Court of
Canada decision in Dukart v. Surrey (District), [1978] 2 S.C.R. 1039 at p.1052-53,
86 D.L.R. (3d) 609 [Dukart cited to S.C.R] as support for its submission
that, where the words of a recital manifest a clear intention, the Courts have
inferred that the parties intended that these words be given effect.
[52]
The case of Dukart does not assist Apotex. Dukart involved
the grant of an “easement” and the question of the true intentions of the
parties. In that case, the body of the agreement contained no language with
respect to the extent of the rights granted under the agreement. The recital
clause was used by the Supreme Court to provide the necessary meaning to the
agreement.
[53]
The case before me is different in that provisions in the body of the MACI
Agreement speak to the intent of the agreement and the scope of the “transfer”
from Merck & Co. to MACI. The use of a recital or preamble as an
interpretative aid must always be approached with caution. As pointed out by
Justice Abella (as she then was) in Lay v. Lay (2000), 47 O.R. (3d) 779,
184 D.L.R. (4th) 652 (Ont. C.A.) at paragraph 12, leave
to appeal to SCC refused, [2000] S.C.C.A. No. 369 (QL), 264 N.R. 398 (note):
There is no doubt that
an introduction or a preamble can provide interpretative assistance, but I see
no basis for accepting the novel proposition that its terms can triumph over
those in the body of the contract.
[54]
In my view, the words of the MACI Agreement establish the creation of a
licence and not a conveyance of all rights in the '380 Patent. The use
of the word “remaining” in the recital does not “triumph over” the words of the
agreement. This is sufficient to defeat the argument of Apotex.
[55]
However, even if I accept that there may be ambiguity in the MACI
Agreement, I am satisfied that the parties to the MACI Agreement did not intend
to convey the entire right, title and interest in the '380 Patent. One indication
of the intent of the parties to an agreement is the behaviour of the parties.
If the MACI Agreement is not clear on its face, it is of assistance to examine
the behaviour of the parties after the execution of the agreement. Was the
behaviour of Merck & Co., from November 1992, consistent with a company who
had given up its entire right, title, estate and interest in the '380 Patent?
Clearly, the answer is “no”. If there had been such intent, why would Merck
& Co. commence and pursue this litigation for 13 years in its own name?
Further, why would Merck & Co. remain as the named patentee on the '380
Patent?
[56]
I am satisfied that the MACI Agreement did not operate as a conveyance of
the entire right, title and interest of Merck & Co. to MACI. Merck &
Co. has standing to bring this action.
V. Claims
Construction
A. Principles
of Claims Construction
[57]
The first step in a patent suit is to construe the claims, in accordance
with principles that are well-established in the jurisprudence (see, for
example, Whirlpool Corp. v. Camco Inc., 2000 SCC
67, [2000] 2
S.C.R. 1067 [Whirlpool]). This jurisprudence teaches that
claims are to be interpreted in a purposive way in order "to achieve
fairness and predictability and to define the limits of the monopoly" (Dimplex
North America Ltd. v. CFM Corp., 2006 FC
586, 292 F.T.R.
38 at para. 49 [Dimplex], aff'd 2007 FCA
278, 60 C.P.R.
(4th) 277).
[58]
Construction of the claims is a matter for the Court to determine. The
Court is called on to determine, on an objective basis, what a hypothetical
skilled person would have understood the invention to mean (Whirlpool,
above, at paras. 45, 53). Where a patent is of a highly technical nature, the
person skilled in the art will be someone possessing a high degree of expert
scientific knowledge in the particular field of art to which the patent relates
(Aventis Pharma Inc. v. Apotex Inc., 2005 FC 1283, 278 F.T.R. 1 [Ramipril
I (FC)]; Apotex Inc. v. Syntex Pharmaceuticals International Ltd et al
(1999), 166 F.T.R. 161 at para. 38, [1999]
F.C.J. No. 548 (QL)(F.C.T.D.).
[59]
Where necessary, the whole of the patent, and not only the claims, should
be interpreted (Eli Lilly Canada Inc. v. Apotex Inc., 2008 FC
142, 63 C.P.R.
(4th) 406 at para. 25; Eli Lilly Canada Inc. v. Novopharm Ltd.,
2007 FC
596, 58 C.P.R.
(4th) 214 at para. 103). The Court should construe the claims in
light of the description in the specification, assisted by experts as to the
meaning of technical terms if such terms cannot be understood by the Court from
reading the specification (Shire Biochem Inc. v. Canada (Minister of Health),
2008 FC
538, 328 F.T.R.
123 at para. 22 [Shire]; Whirlpool, above, at para.
45).
[60]
It is also important to recognize that purposive construction should be
directed at the points in dispute between the parties (Shire, above, at
para. 22).
[61]
Lastly, as the '380 Patent was issued under the old Patent Act, all
claims at issue are to be construed as of the date the patent was granted and
issued (Pfizer Canada Inc. v. Canada (Minister of Health), 2005 FC
1725, 285 F.T.R.
1 at para. 36). For the '380 Patent, that date is January 31, 1984.
[62]
With these overarching principles in mind, I turn to the patent in
question.
B. The
hypothetical skilled person
[63]
As noted, claims must be construed from the view of a hypothetical skilled
person. Thus, as a preliminary matter, I must define what attributes would be
held by our hypothetical skilled person.
[64]
In its final written argument, Apotex described the person to whom the
'380 Patent is addressed as follows:
The skilled addressee
of the ’380 Patent is a notional person having a thorough knowledge of
cultivating fungal micro‑organisms. Such a person may have an advanced
degree, such as a Ph.D., in biochemistry, mycology or industrial biochemical
processes and several years of related practical experience in an industrial
setting. The skilled addressee would also include pharmaceutical formulators, and
medical and organic chemists interested in using the compounds of the alleged
invention to treat hyperlipemia and hypercholesteremia.
[65]
Given the nature of this product-by-process patent, I believe that
experience and knowledge related to fungal micro-organisms is fundamental. This
expertise would, in my view, include both academic qualifications and technical
experience. Identification of micro-organisms, developing, recognizing and
identifying productive strains and growing cultures in appropriate media for
commercialization are all aspects of the '380 Patent with which the skilled
addressee must be familiar. I agree with Dr. Clardy when he states (Clardy
Expert Report, Exhibit 17, para. 25):
… [F]ermenting fungi
to obtain secondary metabolites requires experience beyond ordinary academic
training and because isolating natural products from fermentations requires
the interplay of chemistry, biosynthesis and biological assays beyond formal
academic training.
[66]
Since the invention consists of processes for preparing compounds that are
targeted to lower serum cholesterol, the skilled person would have sufficient
knowledge of medical and organic chemistry to be able to understand cholesterol
biosynthesis. This expertise could be acquired through academic training or
clinical practice.
[67]
With these remarks on some of the skills necessary, I
accept Apotex’s description of the person to whom the '380 Patent is addressed.
C. The Patent
Specification
[68]
I begin with a brief overview of the patent specification.
[69]
The '380 Patent is what is commonly described as a product-by-process
patent. That is, the inventors do not make a specific (or per se) claim
to the compound lovastatin; rather, they claim the product lovastatin and three
other compounds when the compounds are made by the processes described in the
patent. The '380 Patent is entitled “HYPOCHOLESTEREMIC FERMENTATION PRODUCTS
AND PROCESS OF PREPARATION”. As set out in the summary:
This invention relates
to hypocholesteremic products from the cultivation of a microfungus of the
species Aspergillus. More specifically, it relates to compounds of the
formulae:
as well as
pharmaceutically acceptable salts and lower alkyl and substituted alkyl esters
of the carboxylic acids in which the possible substituent is phenyl, dimethylamino
or acetylamino. The invention also relates to a process of cultivating the
microfungus and isolating from the medium a hypocholesteremic compound of the
above structures. These new compounds have excellent properties of inhibiting
cholesterol biosynthesis and are useful against hypercholesteremia and
hyperlipemia.
[70]
I was assisted in understanding the chemical structures of, and
relationship amongst, the four compounds identified in this summary (and set
out in claim 1) by Drs. Lasure, Clardy and Samson.
[71]
Compound I is the marketed product named lovastatin. Compound II is
dihydro-lovastatin. It differs from lovastatin (Compound I) only in relation to
the bonds in the double ring structure. Compounds I and II are both lactones,
meaning that the ring at the top right of the structure is closed. Compound III
is the open acid or hydroxy acid form of Compound I and Compound IV is the open
acid or hydroxy acid form of Compound II. Each of Compounds III and IV has an
open ring at the top with a COOH (carboxyl) group.
[72]
At p. 2 of the patent description, the inventors disclose, as prior art,
the work and patents of Endo and others related to the compound compactin:
Recently, Endo et al.,
described (U.S. 4,049,495 and 3,983,140) a fermentation product obtained by
cultivation of a micro-organism of the genus Penicillium and isolation
from the medium. They called it ML 236 B and determined its structure together
with two related compounds 236 A and 236 C. Its structure, under the name
compactin, was also determined by A.G. Brown, T.C. Smale, T.J. King, J.
Chem. Soc. (Perkin I) 1165 (1975). This compound has been found to be
an inhibitor, in vivo, of the biosynthesis of cholesterol.
[73]
The inventors then distinguish their invention from that of the prior art.
We have found that
unexpectedly, the cultivation of a micro-organism very different from that
employed by Endo, a microfungus of the genus Aspergillus, produces new
substances that are also very potent inhibitors of the biosynthesis of
cholesterol in mammals. We have further found that these substances comprise
principally the new compounds I, II, III and IV, of the above structures,
accompanied by only traces of other compounds. These new compounds are much
more potent inhibitors of cholesterol synthesis in vivo than is the
compound, ML236B described by Endo.
[74]
In short, the patent specification discloses that the inventors of this
patent built on the existing work of Endo and others in relation to the
anti-cholesterol properties of compactin. They discovered that the Compounds I,
II, III and IV, cultivated from “a microfungus of the genus Aspergillus”,
rather than from the genus Penicillium, were more potent inhibitors of
cholesterol synthesis in vivo than compactin.
[75]
At p. 3 of the patent specification, the inventors state that, “The compounds
of this invention are highly useful as antihypercholesteremic agents for the
treatment of atherosclerosis, hyperlipemia and like diseases in humans.”
[76]
Beginning on p. 4, the inventors begin their more detailed description of
how this invention relates to a process for producing the identified compounds.
From the experts, I have learned that the method of production of the '380
Patent is described as a “fermentation”. Dr. Lasure described this as
follows (Lasure Expert Report, Exhibit 48, para. 20):
Unlike more
traditional processes by which chemists may synthesize chemical compounds in a
laboratory or a factory using controlled chemical reactions in vitro, the
process disclosed in the ‘380 Patent is a biological process involving the use
of a specific fungus that synthesizes lovastatin in vivo when that fungus is
grown in or under certain conditions which the ‘380 Patent calls a
“fermentation”.
[77]
The inventors describe their use of two sample micro-organisms from the
culture collection of Merck and Co., referred to as MF-4833 and MF-4845. These
two micro-organisms were placed on deposit with the American Type Culture
Collection (ATCC) and assigned accession numbers ATCC 20541 and ATCC 20542
respectively. It is clear that the inventors are not limiting their invention
to the use of these two micro‑organisms.
Although the use of
these is described in connection with the process of this invention, other organisms
of the genus Aspergillus including mutants of the above ones are also
capable of producing these novel compounds and their use is contemplated in
carrying out the process of this invention.
[78]
In the paragraph that follows, the inventors disclose that:
The morphological
characteristics of the micro-organisms MF‑4833 and MF-4845 have been
found to be those of the genus Aspergillus. Using the criteria specified
in the standard authority
. . . and by
comparison with known species, it has been determined that both strains are Aspergillus
terreus.
[79]
Beginning at the bottom of p. 5, the process of fermentation is described,
with reference to such matters as illustrative media, optimal temperature
ranges and the pH of nutrient media suitable for growing the culture. More
detail is provided regarding fermentation scaling, from the initial culture in
small flasks to the large-scale fermentation tanks. Once the fermentation broth
is made, the reader is instructed that the compounds are conveniently isolated
from the fermentation broth as lactones I and II or, alternatively, as salts of
Compounds III and IV. Other methods of yielding the compounds of the invention
are disclosed.
[80]
The physico-chemical properties of the compounds are set out starting at
p.8. At p. 9-10, the inventors set out their belief, “with a considerable
degree of certainty”, the stereo chemical structures of Compounds I and III.
Similar data and stereo chemical structures of Compounds II and IV are
described at p. 11-12.
[81]
The specification, from p. 13 to 43, illustrates the “invention” with 27
examples.
D. The claims
in issue
[82]
Merck, as it outlined in a response to a demand for
particulars dated July 8, 1998, alleged that claims 1 to 8, 10, 11, 13 to 16,
18 and 19 of the '380 Patent were infringed by the Defendants. Merck
specifically stated that it was not relying upon every claim in the '380
Patent.
[83]
In their Statements of Defence and Counterclaim, Apotex asserts that all
of the claims of the '380 Patent are invalid. Subsequent submissions of the
parties lead me to the conclusion that, at this stage, the only claims still in
issue – and that require construction – are claims 1 to 8 and
13 to 15. I have set out claim 1 in full
below. The remaining claims are included in Appendix B to these reasons.
1. A process
of producing the compounds of structural formulae:
which compromises
fermenting a nutrient medium with a microorganism of the genus Aspergillus
terreus and isolating the products and when desired converting said
products to their corresponding pharmaceutically acceptable salt or lower alkyl
ester or a substituted lower alkyl ester wherein the substituent is phenyl,
dimethylamine or acetylamine or the cation of the salt is derived from ammonia,
ethylenediamine, N-methyl-glucamine, lysine, arginine or ornithine.
[84]
The disagreement between the parties focuses on aspects of claim 1. The
proper construction of the other claims at issue flow from a resolution of the construction
of claim 1. That is, a proper construction of claim 1 will be determinative of
the main points in dispute for the remainder of the claims in issue, claims 2
to 8 and 13 to 15.
[85]
The areas of disagreement between Apotex and Merck are the following:
1.
What is the meaning of the phrase “a micro-organism of genus Aspergillus
terreus” in claim 1?
2.
What is the meaning of the word “isolating” in claim 1?
3.
Does the '380 Patent promise that all strains of Aspergillus terreus
will be capable of producing the four compounds of the invention?
4.
What is the promised use of the claimed invention?
[86]
The final two construction issues relate to the “promise” of the '380
Patent. The question of what is promised – or not – by the '380 Patent is
primarily relevant to the question of the utility of the patent. However, it is
an analysis that logically forms part of the '380 Patent claims construction.
[87]
Generally, ascertaining the promise of a patent is an exercise that
requires the assistance of expert evidence (Bristol-Myers
Squibb Co. v. Apotex Inc., 2007 FCA 379, F.C.J. No. 1579 (QL) at para. 27).
This is because the promise should be properly defined, within the context of
the patent as a whole, through the eyes of a person of skill in the art.
E. The
meaning of “a microfungus of genus Aspergillus terreus” in claim 1
[88]
The first construction issue raised by Apotex relates to the proper
interpretation of the words “a micro-organism of the genus Aspergillus
terreus” in claim 1.
[89]
Claim 1 speaks to a process for producing four compounds “which comprises
fermenting a nutrient medium with a micro-organism of the genus Aspergillus
terreus . . ..” Apotex argues that, on a proper interpretation
of claim 1, a skilled person would read the words “genus Aspergillus terreus”
as referring to all micro-organisms in the genus Aspergillus. Merck
asserts that claim 1 is limited to micro-organisms of the species Aspergillus
terreus.
[90]
The nomenclatures used in the '380 Patent would be understood by any high
school biology student (and even this judge) as part of the binomial system of
naming living organisms. All living
organisms are named according to a
hierarchy of classifications. The hierarchy is as follows:
(i) Kingdom
(ii) Phylum
(iii)
Class
(iv)
Order
(v)
Family
(vi)
Genus
(vii)
Species
In accordance with the accepted binomial
convention, living organisms are identified by using the name of the genus
(capitalized) together with the name of the species within the genus (lower
case).
[91]
There is no such genus as Aspergillus terreus. As I have learned
from the experts in this case, Drs. Clardy, Lasure and Samson, the
well-established rules of taxonomy dictate that Aspergillus is a genus
and that Aspergillus terreus is a species within the genus Aspergillus.
As submitted by the parties, the use of the term “genus Aspergillus terreus”
in claim 1 can mean one of two things:
(a)
the inventors were claiming only those compounds made with micro-organisms
of the species Aspergillus terreus, and inadvertently used the
term “genus” in place of “species” (Merck’s position); or,
(b)
the inventors were claiming compounds produced from any micro-organism
falling within the genus Aspergillus (Apotex’s position).
[92]
Even without expert assistance, it appears to me that the first option
provides the preferable interpretation. There is no question that the skilled
reader would recognize Aspergillus terreus as a species – that is, a
subset of the genus Aspergillus. The skilled reader would assume that
the inventors intended that claim1 include only the micro-organisms of the
genus Aspergillus that belong to the species terreus. To read the
phrase as including all species within the genus Aspergillus would
ignore the plain meaning of the term terreus, as used in the claim.
[93]
Not only does my construction of the words accord with common sense, it is
consistent with the opinions of Drs. Lasure and Clardy. In the view of Dr.
Clardy (Clardy Expert Report, Exhibit 17, para. 33):
… [T]he words Aspergillus
terreus were in January 1984 and remain today words which, by definition,
mean and would be read by the skilled person to describe a subset or
sub-category of Aspergillus that is not and cannot include the entire Aspergillus
genus.
[94]
Initially, Dr. Samson expressed a different interpretation of this phrase
and concludes that (Samson Expert Report, Exhibit 109, para. 37):
… [B]ased on the
repeated references in the patent to the use of a micro-organism of the “genus Aspergillus”
and from the balance of my review of the ’380 Patent described above, it is my
opinion that a person skilled in the are would have concluded that the
inventors did not intend to place any limits on the micro-organisms that can be
used in the process other than that they be from the genus Aspergillus.
[95]
I acknowledge that the disclosure or specification of the '380 Patent
makes a number of references to the “genus Aspergillus”. Nevertheless,
the key problem with Dr. Samson’s interpretation of the phrase “genus Aspergillus
terreus” in claim 1 is that it ignores completely the word “terreus”.
If I were to accept Dr. Samson’s opinion, I would be expanding the claim from “Aspergillus
terreus” to the much broader designation of “Aspergillus”. As Dr.
Samson stated, there are over 250 species that fall within the genus Aspergillus
(Samson Expert Report, Exhibit 109, para. 17).
[96]
Dr. Samson’s approach to claims construction is contrary to the teachings
of the Supreme Court in Whirlpool, above, at paragraph 52, where Justice
Binnie refers to the statement of Taschereau J. in Metalliflex Ltd. v. Rodi
& Wienenberger AG (1960), [1961] S.C.R. 117 (S.C.C.), at p. 122:
The claims, of course,
must be construed with reference to the entire specifications, and the latter
may therefore be considered in order to assist in apprehending and construing a
claim, but the patentee may not be allowed to expand his monopoly
specifically expressed in the claims "by borrowing this or that gloss from
other parts of the specifications". [Emphasis added.]
[97]
During cross-examination, Dr. Samson appeared to have qualified or changed
his opinion. Specifically, he agreed that “the inclusion of the word ‘terreus’
in claim 1 excludes all other species of Aspergillus including niger and nidulus
and oryzae and the other 246 [species].”
[98]
Moreover, when read in its entirety, the specification is consistent with
the limitation of the invention to micro-organisms of the species Aspergillus
terreus. For example, the inventors disclose the use of micro-organisms
MF-4833 and MF-4845; these are examples of Aspergillus terreus and not
of some other species within the genus Aspergillus.
[99]
In summary on this point, the words of claim 1 make it very clear that the
patentee is not claiming compounds made with any of the 250 species of Aspergillus;
rather, the boundary of the invention, as claimed, includes the four identified
compounds when made with a single species – Aspergillus terreus. The use
of the word “genus” before “Aspergillus terreus” may have been a simple
inconsequential error by the drafters or the patentee may have intended the
word “genus” to modify only the word “Aspergillus” and not the entire
phrase “Aspergillus terreus”. Regardless, given the specificity of the
term “Aspergillus terreus”, the use of the word “genus” would not change
the meaning ascribed to the phrase by the skilled addressee.
F. The
meaning of “isolating the products”
[100]
The second construction issue concerns that part of claim 1 which states
that the process of producing the compounds “comprises fermenting a nutrient
medium with a micro-organism of the genus Aspergillus terreus and isolating
the products . . .”.
[101]
The parties disagree on the meaning of the phrase “isolating the
products”. Apotex submits that claim 1 requires the production of all four of
the compounds, that claim 2 requires the production of both Compound I and II
(the lactones) and that claim 5 requires the production of both Compound III
and IV (the hydroxy acids). In other words, Apotex argues that, for purposes of
the claims, the compounds must be separated from each other, purified and
crystallized before they are “isolated”. Merck submits that “isolating” simply
means separating the compounds from the fermentation broth and does not require
the compounds to be purified or crystallized.
[102]
The term “isolating” is not defined in the patent. Therefore, it is
necessary to review the specification to determine what meaning was reasonably
intended by the inventors.
[103]
Example 1 of the '380 Patent is entitled “Preparation of Compounds I and
III”. The inventors first set out a procedure for fermenting a particular
culture of Aspergillus terreus. At the end of this step, the skilled
person would have a fermentation broth that is “set aside for isolation of the
product”. After the fermentation is completed, the next step is the “Isolation
of Compound I”. This involves separating the broth solids from the broth
liquids, extracting the liquids using a mixture of solvents and extracting the
solids. The resulting extracts are combined and concentrated to 15 ml of crude
extract. There is no purification or crystallization described as part of the
“Isolation of Compound I”. Further, the next step – “Testing of Compound I” –
is carried out on the crude extract. There is no further purification or
crystallization carried out before the testing.
[104]
In reviewing the examples of the patent, I note that Examples 3, 4 and 5
all refer to isolation without any purification or crystallization.
[105]
From the specification, I conclude that the inventors meant the term
“isolating” to simply refer to separating the compounds from the broth. This
was the interpretation given to the term “isolating” by Dr. Clardy who opined
that (Clardy Expert Report, Exhibit 17, para. 102):
In the ‘380 Patent
"isolating" does not necessarily require complete separation or
purification of the active compounds. The concept of "isolating the
products" of a fermentation as those words are used in the patent requires
getting the products out of the fermentation broth. The products do not
necessarily have to be isolated from one another nor be crystalline, nor be
completely purified. All of this would be understood by the skilled person
reading the patent in January 1984. I note that the "Isolation of Compound
I" in Example 6 is more complex and includes the further purification and
crystallization of a specific fraction containing Compound I, but example 1
makes clear that such steps are optional and not necessarily required for
"isolating" as that word is used in the patent.
[106]
Dr. Samson provided a contrary view. In his Reply Expert Report, Dr.
Samson opines as follows (Samson Reply Expert Report, Exhibit 11, para. 41):
The '380 Patent says that
the compounds are extracted or isolated from the fermentation broth as
“hypocholesteremic compounds” (see page 2, lines 4 to 7). This would have been
understood by a person skilled in the art to mean that the invention requires
that the compounds be removed from the fermentation broth and isolated from any
other compound in the broth, and then purified and crystallized so that they
can be useful as “hypocholesteremic compounds”.
[107]
I have difficulty with Dr. Samson’s understanding of the term “isolating”.
Foremost, this interpretation ignores much of the content of the specification
that describes the testing of the crude extract. During cross-examination, Dr.
Samson acknowledged that, in some of the examples, there was no purification,
separation or crystallization prior to testing. Nevertheless, he clung –
unreasonable, in my view – to the opinion that the skilled person would presume
that “isolation” or “isolating” includes the steps of extraction,
crystallization and separation.
[108]
Beyond the disclosure of the '380 Patent, the interpretation proposed by
Merck is also supported by the reading of claim 1 in the context of the other
claims – in particular claim 13. The general process is set out in claim 1.
Claim 13 is a claim to a compound selected from Compounds I, II, III and IV.
Had the inventors intended claim 1 to require a separation of each compound
from the others, they could have used similar language of selection. The use of
the phrase “isolating the products” rather that “isolating each
product” is a strong indication, for the skilled reader, that the inventors
did not intend that each of the compounds be separated from each other,
purified and crystallized.
[109]
In sum, I am satisfied that, on a proper claims construction, the words
“isolating the products” in claim 1 do not require that the relevant compounds
be separated from each other, purified or crystallized prior to testing.
G. Inclusion
of non-producing strains
[110]
For the third construction question, Apotex submits that, whether the
Court construes the phrase “genus Aspergillus terreus” in claim 1 to
include all micro-organisms or fungi within the genus Aspergillus or
just those within the species Aspergillus terreus, the '380 Patent promises that all such micro-organisms can be used to
produce the compounds in the Patent. Merck, on the other hand, asserts that the
person skilled in the art would know – and eliminate from coverage of the '380
Patent – any strain or fungus that cannot produce the claimed compounds.
[111]
Neither claim 1 nor the specification explicitly states that the '380
Patent excludes non-producing strains of Aspergillus terreus. The
question to be determined is whether the skilled addressee, in 1984, would know
that the claims of the '380 Patent are limited to the producing strains of Aspergillus
terreus.
[112]
An essential element of the invention embodied in the '380 Patent is the
production of particular compounds through the process of fermentation or
cultivation of fungi. As described by Dr. Lasure, who has extensive experience
working with such micro-organisms, “a culture of Aspergillus terreus is
a living sample of a fungus from that species” (Lasure Expert Report, Exhibit
48, para. 60). Thus, within species, there are variations.
[113]
Dr. Sorensen described the organic compounds that are produced by fungi as
“secondary metabolites”. “Secondary metabolites” are compounds produced as a
result of the metabolic function of the initial fungal micro-organism during
the fermentation process (Sorensen Expert Report, Exhibit 132, para. 6). While
Dr. Clardy opined that “in the general case it was expected that a particular
isolate from a producing species would be expected to produce a given
metabolite,” he cautioned that there is always a possibility of non-production,
for a number of reasons, all of which would have been known in 1984 (Clardy
Expert Report, Exhibit 17, paras. 39-41):
·
fungi can be intentionally mutated to disrupt the genes responsible for
making a metabolite;
·
fungi can lose a producing ability over time because of subculturing, or
mishandling, or reasons that are never understood;
·
some fungi in a producing species simply do not have production
capability, although Dr. Clardy thought that this would be unusual except where
the isolates have been maintained by serial subculturing;
·
fungi tend to change randomly, especially when stored and maintained
artificially by scientists;
·
if during subculturing, a sample of a variant is taken for further growth,
and especially if there is repeated subculturing, the fungus can become one in
which a characteristic of the parent culture is lost; and
·
physical deterioration of the fungus will lead to loss of specific
metabolic functions.
[114]
Dr. Clardy summed up the situation as follows (Clardy Expert Report,
Exhibit 17, para. 42):
It was part of the
common knowledge of the skilled person in 1984 and such a person would have
known, with virtual certainty, that among the many isolates (or strains) of a
given species there would inevitably be found isolates that have lost the
capacity to produce a particular metabolite under particular conditions, or in
some rare cases, that never had an ability to produce the metabolite at all.
[115]
Dr. Samson did not share this opinion. In his view, the skilled person
would interpret the claims as including the production by Aspergillus of
all four identified compounds (Samson Expert Report, Exhibit 110, para.
52). He disagreed that the exclusion of non-producing strains was implicit.
However, during cross-examination, he acknowledged that the opinions of Drs.
Lasure and Clardy were “both scientific opinions that a reasonable person of
ordinary skill in the art could have reached in January 1984”.
[116]
Dr. Samson, during cross-examination, also agreed that the person of
ordinary skill in 1984 would have been fully aware that not every strain in a
species will make a given metabolite. Even more specifically, Dr. Samson
acknowledged that, in 1984, a skilled person reading the claims of the '380
Patent would know that there are strains of Aspergillus terreus that
would not produce lovastatin.
[117]
In sum, I prefer the opinions of both Drs. Lasure and Clardy to the effect
that a skilled person would know, from his or her general knowledge, that:
(a)
there are many variations in the micro-organisms within the species Aspergillus
terreus, such that not every micro-organism within the species will
necessarily provide the desired results and that some testing and routine
experimentation will be required; and
(b)
the term “nutrient medium”, as used within the patent description, would
include the media used in the examples and other media that, upon routine
testing, would result in the desired compounds.
[118]
In light of this general knowledge, Drs. Clardy and Lasure were both of
the opinion that the skilled person would recognize – even without an explicit
limitation in the claims – that the claims of the '380 Patent exclude
non-producing strains of Aspergillus terreus. The use of the word
“producing” in claim 1 tells the skilled person that non-producing strains are
excluded, even though the explicit words are not used.
[119]
Further support for the conclusion reached by Drs. Clardy and Lasure is
contained within the disclosure. In addition to the disclosure of the structure
of the compounds and the therapeutic activity of the compounds, the skilled
person is provided with examples of the media and conditions that could be
used.
[120]
Moreover, the skilled person would also bring to his or her laboratory “bench”,
a set of skills used routinely. In light of the nature of fungi and their use
in the pharmaceutical industry, it appears logical to me that the skilled
person would have extensive experience with the types of experimentation and
testing that are used to identify and optimize producing micro-organisms.
During cross‑examination, Dr. Samson confirmed that a number of
experiments, known in 1984, could have been performed simultaneously, in a
short period of time, to identify the producing micro-organisms. He agreed that
about 300 shake flask experiments could be run at one time. The skilled person
could rapidly screen a large numbers of isolates of Aspergillus terreus
to determine which strains are producing. Moreover, since, on my construction,
the claims are limited to strains of the species Aspergillus terreus,
there are manageable boundaries on the testing that would be required.
[121]
Having considered the evidence of the three experts, I am persuaded that
the opinions of Drs. Lasure and Clardy on this point are to be preferred to
that of Dr. Samson. An implicit requirement that non-producing strains are
excluded from the coverage of claim 1 is “being neither benevolent nor harsh
but rather seeking a construction which is reasonable and fair to both patentee
and public” (Consolboard Inc. v. MacMillan Bloedel (Saskatchewan) Ltd.,
[1981] 1 S.C.R. 504 at p.520, 56 C.P.R. (2d) 145 [Consolboard cited to
S.C.R.]). I do not accept Apotex’s assertion that the '380 Patent states,
implies or promises that all strains of Aspergillus terreus will be
capable of producing the four compounds of the claimed invention.
H. The
promised use of the '380 Patent
[122]
As a final construction issue, Apotex submits that the '380 Patent makes
the explicit promise that the four compounds identified in claim 1 are “highly
useful as anti-hypercholesteremic agents for the treatment of atherosclerosis,
hyperlipemia and like diseases in humans”.
[123]
The “promise” of the '380 Patent appears to be clearly set out in at least
two places in the specification. In the “Summary of the Invention”, at p. 2 of
the patent, the patentees state that:
These new compounds
have excellent properties of inhibiting cholesterol biosynthesis and are useful
against hypercholesteremia and hyperlipemia.
[124]
A slightly more detailed promise is found at p. 3, where the patentees
explain that:
The compounds of this
invention are highly useful as antihypercholesteremic agents for the treatment
of atherosclerosis, hyperlipemia and like diseases in humans.
[125]
Dr. Samson’s opinion is that the detailed statement in the patent (p.3) expresses
the promise of the '380 Patent (Samson Expert Report, Exhibit 110, para. 25).
Although Dr. Lasure did not directly respond to the question of what was the
promise of the patent, she described the uses of the '380 Patent to include the
following (Lasure Expert Report, Exhibit 48, para. 21):
With respect to uses,
the ‘380 Patent discloses that the four compounds (and salts and esters of
them) can be used to inhibit cholesterol biosynthesis, can be used against
hypercholesteremia (high levels of cholesterol in the blood) and hyperlipidemia
(high levels of lipids in the blood) and can be used as antifungal agents (to
kill or inhibit growth of fungi on plants).
Anti-fungal properties have not been
referred to by Apotex in this matter. The issue for this trial is focused on
the medical use.
[126]
Based on the words of the specification and supported by the opinions of
Drs. Lasure and Samson, I find that the skilled person would read the '380
Patent as promising that the compounds (or secondary metabolites) produced from
strains of Aspergillus terreus, by the fermentation process identified
in the patent, are “useful as antihypercholesteremic agents for the treatment
of atherosclerosis, hyperlipemia and like diseases in humans”.
I. Summary on Claims
Construction
[127]
Considering the words of the claims of the '380 Patent and the
specification and guided by the expert testimony, the relevant claims of the
patent should be construed in the following manner:
·
Claim 1 is a claim to a process for producing the four identified
compounds by a fermentation process using the range of nutrient media and
conditions described in the specification (or as would be generally known to
the skilled person), following which the compounds are isolated or extracted from
the fermentation broth by any of
the known
means identified in the specification (or as would be generally known to the
skilled person). Of particular relevance to this litigation:
○
the micro-organism or fungus to be used is a strain of the species Aspergillus
terreus, excluding those strains that are unable to produce the desired
compounds and excluding micro-organisms from other species within the genus Aspergillus;
and
○
after the fermentation stages of the process, the resulting broth may
contain any or all of the four compounds.
[128]
With this construction of claim 1, the construction of claims 2 to 8
follows. Each of these claims is a “subset” of claim 1, whereby the claim is
restricted to:
·
the process of producing only Compounds I and II (claim 2) or Compounds
III and IV (claim 5);
·
the process of producing the identified compounds using a particular
originating micro-organism (claim 3 and 6); and
·
the process of producing the identified compounds using certain
operational requirements (claims 4, 7 and 8).
[129]
Claim 13 claims any one of the four identified compounds (the same
compounds as described in claim 1) when made by the process of claim 1 “or by
an obvious chemical equivalent”. Claim 14 is a similar claim to either Compound
I or II when made by the process of claim 2 “or by an obvious chemical
equivalent”. Claim 15 claims each of Compound III and IV when made by the
process of claim 5 “or by an obvious chemical equivalent”.
[130]
Finally, I find that the skilled person would read the '380 Patent as promising
that the compounds (or secondary metabolites) produced from strains of Aspergillus
terreus, by the fermentation process identified in the patent, are “useful
as antihypercholesteremic agents for the treatment of atherosclerosis,
hyperlipemia and like diseases in humans”.
VI. Infringement
– Background
A. Introduction
[131]
Having established the proper construction of the relevant claims of the
'380 Patent, I now turn to the question of infringement.
[132]
Section 44 of the Patent Act confers on a patentee and his legal
representatives "the exclusive right, privilege and liberty of making,
constructing, using and vending to others to be used the invention" of a
patent. Merck claims that the Defendants infringed their rights under the '380
Patent by the production of lovastatin
using the Aspergillus terreus micro-organism. Specifically, Merck claims
infringement in three different scenarios:
1.
infringement through the manufacture (during Phase 1 of production
described above), by AFI in Winnipeg, of quantities of lovastatin included in batch CR0157;
2.
infringement, between April 1997 and March 1998, through the manufacture
by Blue Treasure of quantities of infringing lovastatin that were shipped to
AFI, when Blue Treasure was allegedly “salting” the lovastatin shipments with
infringing Aspergillus terreus lovastatin; and
3.
infringement from March 1998, when Blue Treasure was allegedly shipping
lovastatin manufactured with Aspergillus terreus.
[133]
Subject to the possible exceptions of regulatory or experimental use, the
making, constructing, using or vending lovastatin using Aspergillus terreus
would be an infringement of the '380 Patent. Thus, if Merck can satisfy the
Court that certain volumes of lovastatin made from the product received from
Blue Treasure were manufactured from or contained, through “salting”, Aspergillus
terreus lovastatin, infringement has been established. Similarly, if Merck
can persuade the Court that batch CR0157 contained lovastatin manufactured from
Aspergillus terreus, infringement has been proved.
B. Burden
[134]
The first point to be made is that proof of infringement is subject to the
civil standard of proof. Merck’s burden – whatever it may be – is met if
infringement can be shown on a balance of probabilities. Stated in different
words, Merck will succeed if it is more likely than not that infringement
occurred.
[135]
It is trite law that the party alleging infringement bears the burden of
proving infringement (see Monsanto Canada Inc. v. Schmeiser, 2004 SCC
34, [2004] 1 S.C.R. 902 at para. 29 [Monsanto]). However, consideration
must be given to the scheme of the Patent Act and, in particular, to s.
39(2). Under the Patent Act applicable to this action, s. 39(1) provides
that:
|
39. Naturally occurring substances intended for food
or medicine—(1) In the case of inventions relating to naturally occurring
substances prepared or produced by, or significantly derived from,
microbiological processes and intended for food or medicine, the
specification shall not include claims for the resulting food or medicine
itself, except when prepared or produced by or significantly derived from the
methods or processes of manufacture particularly described and claimed.
[1987, c. 41, s. 14]
|
39. Procédés Microbiologiques Naturels—(1) Lorsqu'il s'agit d'inventions couvrant des substances que l'on trouve dans la nature, préparées ou produites, totalement ou pour une part notable, selon des procédés microbiologiques et destinées à l'alimentation ou à la médication, aucune revendication pour l'aliment ou le médicament ne doit être faite dans le mémoire descriptif, sauf pour celui ainsi préparé ou produit selon les modes du procédé de fabrication décrits en détail et revendiqués.
[1987, ch. 41, art. 14 ]
|
[136]
This provision is followed by s. 39(2) which states that:
|
(2) In an action for infringement of a patent
where the invention relates to the production of a new substance, any
substance of the same chemical composition and constitution shall, in the
absence of proof to the contrary, be deemed to have been produced by the
patented process.
|
(2) Dans une action en contrefaçon de
brevet où l’invention porte sur la production d’une substance nouvelle, toute
substance formée des mêmes composants et éléments chimiques est, en l’absence
de preuve contraire, réputée avoir été produite par la procédé breveté.
|
[137]
On its face, s. 39(2) applies to the facts before me. In Merck’s opinion,
the '380 Patent is to an invention that relates to the production of lovastatin
– a “new substance”. The lovastatin produced in any of Phases 1, 2 or 3 of the
Defendants’ manufacturing is a substance with the same chemical composition and
constitution as that produced by the process of the '380 Patent. As such, s.
39(2) would apply and, absent proof to the contrary, such lovastatin would be
deemed to be produced by the process of the '380 Patent. With respect to
lovastatin manufactured as part of Phase 1 (except for batch CR0157), Merck
accepts that there is “proof to the contrary”. However, Merck asserts that, for
lovastatin that is contained in batch CR0157 and all production sourced from
Blue Treasure, s. 39(2) applies and the production must be deemed to be made
from Aspergillus terreus, thereby infringing the '380 Patent.
[138]
The dispute between the parties centres on the meaning of the words “new
substance” in s. 39(2).
[139]
Merck argues that the substances (Compounds I-IV) claimed in the '380
Patent are new and novel and s. 39(2) is engaged. While there is no definition
of “new”, the word appears in s. 2 under the definition of “invention”. As
such, for patent purposes, “new” could mean novelty, or a product that has not
been anticipated.
[140]
Apotex submits that the definition of newness has to be determined in
light of patent legislation as a whole. Where a word has a meaning in one
section, it ought to be the same in every section within a document, absent
legislative intent to show that the word can have various meanings. In line
with this argument, Apotex’s counsel, in final argument, acknowledged that ss.
2 and 39(2) use the word “new”. However, the word does not appear in provisions
that deal specifically with novelty (anticipation): ss. 61, 27, 43. Thus, one
cannot say that “new” equates with “anticipation”.
[141]
According to Apotex, the interpretation of “new” must fit into the context
of s. 39(2) and its commonsense purpose. In oral submissions, Apotex argued
that the purpose of s. 39(2) (and its presumption of infringement) was:
[…] to deal with the
impossibility of a plaintiff, when it comes to a process in a
product-by-process claiming form, not being able, absent proof from the
defendant of what that process is, to challenge the infringing nature of that
process.
[142]
In line with this purpose, once another process is disclosed for the same
product, the “newness” of the substance per se no longer exists. Apotex also
asserted that newness can be lost in a number of other ways: prior
commercialization, disclosure and use of the substance.
[143]
On the facts of this case, Apotex notes that the application that resulted
in Canadian Patent No. 1,129,794 (the '794 Patent
or the Endo Patent) related to the claims for lovastatin and was filed in Canada before the
'380 Patent. The '794 Patent also had an
earlier priority and issue date. It publicly describes an alternate process to
create lovastatin. In Apotex’s view, the same can be said for lovastatin
created from Red Yeast Rice.
[144]
As argued by the parties, there is no direct case law on the
interpretation of “new” in s. 39(2). It is helpful to return to first
principles of statutory interpretation.
[145]
The starting point of my analysis is the general principle clearly stated
by the Supreme Court in Rizzo & Rizzo Shoes Ltd., Re, [1998] 1
S.C.R. 27 at paragraph 21, 154 D.L.R. (4th) 193 (See also Bell
ExpressVu Ltd. Partnership v. Rex, 2002 SCC 42, 2 S.C.R. 559 at para. 26,
and cases cited therein):
[…] Elmer Driedger in
Construction of Statutes (2nd ed. 1983) best encapsulates the approach upon
which I prefer to rely. He recognizes that statutory interpretation cannot be
founded on the wording of the legislation alone. At p. 87 he states:
Today there is only
one principle or approach, namely, the words of an Act are to be read in their
entire context and in their grammatical and ordinary sense harmoniously with
the scheme of the Act, the object of the Act, and the intention of Parliament.
[146]
Sullivan on the Construction of Statutes, 5th ed. (Markham, Ont.: LexisNexis, 2008) (Sullivan)
comments on the modern principles as articulated by Driedger (p. 3):
The court must adopt
an interpretation that is appropriate. An appropriate interpretation is one
that can be justified in terms of (a) its plausibility, that is, its compliance
with the legislative text; (b) its efficacy, that is, its promotion of
legislative intent; and (c) its acceptability, that is, the outcome complies
with accepted legal norms; it is reasonable and just.
[147]
Furthermore, in relation to the textual analysis of legislation, there are
a number of relevant principles: (a) the presumption of consistent expression (Sullivan,
above, pp. 214-23), and (b) the presumption of coherence (Sullivan,
above, pp. 223-25).
[148]
Under the principle of consistent expression, it is presumed that the
legislature uses language carefully and consistently within the same statute.
As such, same words presumptively have the same meanings. On the flip side, one
can infer, from the use of different words or a different form of expression, that
a different meaning was intended by drafters. This principle was highlighted by
the Federal Court of Appeal in Peach Hill Management Ltd. v. Her Majesty the
Queen (2000), 257 N.R. 193 at paragraph 12, G.S.T.C. 45: “When an Act uses
different words in relation to the same subject such a choice by Parliament
must be considered intentional and indicative of a change in meaning or a
different meaning.”
[149]
According to Sullivan (pp. 221-22), the strength of this
presumption varies. Highly technical statutes and terms that play a key role in
the legislative scheme are strongly presumed to have the same meaning
throughout. For example the definition of “income” in taxation legislation was
considered to be a key term in Mattabi Mines Ltd. v. Ontario (Minister of
Revenue), [1988] 2 S.C.R. 175, 53 D.L.R. (4th) 656. The
presumption of consistent expression is also strong when the repeated words
contribute to a noticeable pattern.
[150]
This presumption, however, can be weakened when one examines the context
surrounding the words: “Identical words may not have identical meanings once
they are placed in different contexts and used for different purposes. This is
particularly true of general or abstract words” (see Sullivan, p. 222; Jevco
Insurance Co. v. Pilot Insurance Co. (2000), 49 O.R. (3d) 760, O.J. No. 2259
(QL) (Ont. Sup. Ct.)).
[151]
The other relevant principle is the presumption of coherence. Here, one
presumes that provisions of the same legislation are meant to work together
logically as parts of a functioning whole.
The parts are presumed
to fit together logically to form a rational, internally consistent framework;
and because the framework has a purpose, the parts are also presumed to work
together dynamically, each contributing something toward accomplishing the
intended goal. […] It is presumed that the body of legislation enacted by a
legislature does not contain contradictions or inconsistencies, that each
provision is capable of operating without coming into conflict with any other (Sullivan,
above, p. 223).
[152]
In applying these principles, the question is: does “new” in s. 39(2) mean
“new” in the ordinary sense, or in the sense of novelty? In other words, to
displace the application of s. 39(2), does Merck have to prove that the
substance of the product-by-process claim in the '380 Patent was novel, or simply
that it was not known before?
[153]
For the reasons that follow, I interpret the word “new” to simply mean a
substance that was not previously known or used, rather than novelty.
[154]
First, within the context of ss. 2 and 39(2), “new” has been used as an
adjective. According to the Gage Canadian Dictionary (W.S. Avis et al.
(ed) (1983), Gage Educational Publishing Co., Toronto), at p. 766, “new” is
defined as “not existing before”. Black’s Law Dictionary (6th
ed.) (St. Paul, Minn.: West Publishing Co, 1990), at p. 1042 describes “new” as
follows:
[…] this word may
denote novelty, or the condition of being previously unknown or of recent
or fresh origin, but ordinarily it is a purely relative term and is
employed in contrasting the date, origin, or character of one thing with the
corresponding attributes of another thing of the same kind or class.
In order to be “new”,
as the word is used in the patent laws, the achievement must be either one that
produces an unusual or improved or advanced result, which was unknown to the
same prior art at the time of the claimed invention; or the achievement
must be one that produced an old result in an unusual and substantially more
efficient, or economical way. [Emphasis added.]
[155]
As seen above, the ordinary meaning of “new” can equate to novelty or
simply a condition of being previously unknown. It is a word that is “purely
relative” in nature.
[156]
Second, I turn to the contextual meaning of the word “new” within the Patent
Act. While there is no dispute that patent legislation is highly technical,
does the word “new” carry a specific and technical meaning? Is it used in a way
that creates a noticeable pattern? Is there a presumption of consistency? My
answer to these questions is “no”.
[157]
“New” in ss. 2 and 39(2) is relative as well; in both instances, the word
is used as an adjective to describe different things. Under s. 2, “new”
describes how an “art” or “improvement” can rise to the level of an invention. The
notion of novelty is part and parcel of this interpretation.
[158]
On the other hand, in s. 39(2), legislators are not describing what
constitutes an invention. This provision relates solely to infringement and
novelty is not directly at issue. Further, the word “new” is not employed to
determine if an “art” or “improvement” constitutes an invention. It merely
describes a “substance” or a “product” in a product-by-process claim. There is
no requirement that the “substance” be inventive. Section 39(2) deals with the
claims in a product-by-process patent. In such claims, the substance cannot be
divorced from its process (see s. 39(1) of the Patent Act). Accordingly,
in a product-by-process claim, whether the substance is novel is not
determinative. It is the process of producing the substance that must be novel,
new, and inventive. In the context of a product‑by-process claim, “new”
does not necessitate a novel (not-anticipated) substance.
[159]
Third, this interpretation is consistent with the rest of the legislative
scheme and avoids internal inconsistencies. If “new” described a novel substance
or medicine as the invention per se, it would contradict s. 39(1) of the
Act.
[160]
Applying this to the case at hand, the claimed invention is not
lovastatin, but lovastatin created through the process of fermenting the
organism Aspergillus terreus. It is clear from s. 39(1) of the Act that
the patentee cannot merely claim lovastatin, a medicinal substance, as the
invention. Lovastatin is merely the product of the product-by-process claim in
the '380 Patent.
[161]
Fourth, ss. 27, 43, and 61, which relate to questions of novelty, have no
mention of the word new.
[162]
Fifth, the jurisprudence supports the interpretation of “new” which means
not previously known. According to Justice Nadon, in Eli Lilly and Co. v.
Nu-Pharm Inc.(1994), 54 C.P.R. (3d) 145 at para. 32, [1994] F.C.J. No.
225 (QL) (F.C.T.D.)[Eli Lilly and Co. v. Nu-Pharm Inc. cited to C.P.R.],
the presumption of infringement does not arise until the plaintiff has
satisfied a minimum evidentiary burden that the substances in question are “new
substances”. This means that the initial burden is on Merck to show that the
substances created from the process in the ‘380 Patent are “new substances”.
The burden is not on Apotex to show that the substances are not new or are
anticipated. Since the burden is on Merck to show that the substances created
from the process in the ‘380 Patent are “new substances”, to equate “new” with
the test of “novelty” would lead to an illogical result. Under the scheme of
Canadian patent law, defendants to an infringement action have the burden to prove
anticipation (a subset of invalidity). Thus, to have the patentee prove novelty
under s. 39(2) would contradict fundamental principles of patent law.
[163]
In sum, “new” is a highly relative term and its definition is dependent on
its context. Within the context of s. 39(2) of the Act, “new substance”
means a substance that was not previously known.
[164]
With respect to the '380 Patent, the question is whether Compounds I-IV
were previously known, or are they simply “old” products? In other words, even
if anticipation is not established by the lovastatin produced by Red Yeast Rice
or by the Endo Patent, can lovastatin be said to have been sufficiently “known”,
and thus, successfully undermine the “newness” of the substance in the '380 Patent?
The answer to these questions is found in the claims construction.
[165]
Claim 1 of the '380 Patent clearly states that it is “a process of
producing the compounds of the structural formulae [Compounds I-IV]”. While the
parties acknowledge that Compound I, lovastatin, is identical to the structure
in Dr. Endo’s ‘794 Patent, claim 1 does not solely fence out Compound I, but
also II, III, and IV. The Defendants presented no evidence that Dr. Endo knew
of, or used, Compounds II, III, and IV. Further, there is no evidence that
lovastatin produced by Red Yeast Rice contains Compounds II, III and IV.
[166]
Accordingly, I agree with Merck’s argument that the combination of
substances produced (Compounds I-IV) is new. There is no evidence that either
the Endo Patent or the fermentation of traditional Red Yeast Rice would produce
such a combination.
[167]
In light of the above, I conclude that s. 39(2) is engaged.
[168]
The question that follows is this: If s. 39(2) applies, what is the
interpretation of “in the absence of proof to the contrary”? Does it mean that Apotex
has an evidentiary burden and that, once this burden is met, the persuasive
burden returns to the Plaintiffs to establish infringement? Or does it mean
that the persuasive burden is established and therefore Apotex must disprove
infringement?
[169]
Merck argues that s. 39(2) mandates that the persuasive burden of proof remains
with Apotex to show non-infringement. This is supported by the language in the
provision – infringement is deemed “in the absence of proof to the contrary”.
According to Merck, this is different from the presumption of validity, which
is weakly worded as merely “in the absence of evidence to the contrary”. Merck
relies on Apotex Inc. v. Tanabe (1994), 59 C.P.R. (3d) 38 at paragraph
92, [1994] O.J. No. 2613 (QL) (Ont. Gen. Div.) [Apotex v. Tanabe cited
to C.P.R.] to interpret the language in s. 39(2) as creating an onus that it “is
not simply an obligation to adduce some evidence to the contrary”. Rather,
“proof to the contrary” is a “much higher onus than the onus simply to adduce
some evidence to the contrary” (Apotex v. Tanabe, above, at para. 92).
Thus, Merck concludes that s. 39(2) is a provision that deems, rather than
presumes, infringement. Consequently, in Merck’s view, the persuasive burden is
on Apotex to prove non-infringement. I disagree.
[170]
Section 39(2) of the Act creates a presumption of infringement that
confers on Apotex an evidentiary burden to rebut infringement. There is no
support for Merck’s argument that s. 39(2) deems infringement and puts the
persuasive burden on Apotex.
[171]
Hughes & Woodley on Patents, 2nd ed.,
looseleaf (Markham, ON: LexisNexis Canada, 2005) at paragraph 45 states that s. 39(2)
creates a presumption of infringement:
This provision creates
a presumption of infringement, but, only once an applicant has satisfied
a minimum evidentiary burden, establishing that the substances in question are
“new” and identical. These provisions, however, may permit the Court to give a
broader interpretation of the claims than a mere purposive construction. This
is to be contrasted with section 45 of the Patent Act which provides for a
presumption of validity in the absence of “evidence” to the contrary.
Speculation, as to alternative processes that may have been used to produce the
product, falls short of the required evidence to the contrary. [Emphasis
added; see also Eli Lilly and Co. v. Nu-Pharm Inc., above.]
[172]
The text notes that, compared to the validity presumption (rebuttable on
an evidentiary basis), the presumption of infringement is stronger. This can be
supported by Justice Campbell’s decision in Apotex v. Tanabe (above, at
para. 92). While Justice Campbell held that “proof to the contrary”
necessitates a higher standard than “evidence to the contrary”, he does not go
so far as to say that “proof to the contrary” equates to a persuasive burden to
prove a fact on a balance of probabilities. As such, comparing the presumptions
of validity and infringement, there is only a difference in degree rather than
nature of the burden.
[173]
Justice Gibson, in Abbott Laboratories v. Canada (Minister
of Health), 2004 FC 1349, 260 F.T.R. 276 at para. 101 [Abbott
Laboratories], examined the words “in the absence of proof to the contrary”
in s. 6(6) of the Regulations:
(6) For the purposes
of an application referred to in subsection (1), if a second person has made an
allegation under subparagraph 5(1)(b)(iv) or (2)(b)(iv) in
respect of a patent and the patent was granted for the medicinal ingredient
when prepared or produced by the methods or processes of manufacture
particularly described and claimed in the patent, or by their obvious chemical
equivalents, it shall be considered that the drug proposed to be produced by
the second person is, in the absence of proof to the contrary, prepared or
produced by those methods or processes. [Emphasis added.]
[174]
According to Justice Gibson, this provision deals with product-by-process
claims, and also creates a presumption that the patent will be infringed.
Despite the strong words of “in the absence of proof to the contrary”, Justice
Gibson was clear that there is no shift in the persuasive burden (Abbott
Laboratories, above, above, at para. 101).
[175]
The Supreme Court, in Circle Film Enterprises Inc. v. Canadian
Broadcasting Corp., [1959] S.C.R. 602 at p.604, 31 C.P.R. 57 [Circle
Film cited to S.C.R], considered similar words in s. 20(3)(b) of the Copyright
Act, R.S.C. 1927, c. 532: “The author of the work shall, unless the
contrary is proved, be
presumed to be the owner of the copyright.” In seeking to characterize the
words in s. 20(3)(b), Justice Judson stated (Circle Film, above, at p.
606):
I take the operation
of a presumption of this kind to be as stated by Wigmore on Evidence, 3rd ed.,
s. 2491(2):
It must be kept in
mind that the peculiar effect of a presumption "of law" (that is, the
real presumption) is merely to invoke a rule of law compelling the jury to
reach the conclusion in the absence of evidence to the contrary from the
opponent. If the opponent does offer evidence to the contrary (sufficient to
satisfy the judge's requirement of some evidence), the presumption disappears
as a rule of law, and the case is in the jury's hands free from any rule.
[Emphasis added.]
[176]
In sum, I conclude that the phrase “in the absence of proof to the
contrary”, in s. 39(2), amounts to an evidentiary burden to rebut the
presumption of infringement.
[177]
Following this, a number of questions are raised: what is an evidentiary
burden? Also, what do the Defendants in this case have to prove in order to
meet their evidentiary burden and rebut the presumption of infringement?
[178]
On the definition of evidentiary burden, the Federal Court of Appeal in Hoffmann-La
Roche Ltd. v. Canada (Minister of Health and Welfare) (1996), 70
C.P.R. (3d) 206 at para. 8, 205 N.R. 331 [Hoffmann-La Roche cited to
C.P.R.] stated:
[…] the
"persuasive burden" or the "legal burden", is the burden of
establishing a case to the civil standard of proof. By contrast, the "evidential
burden" consists of the burden of putting an issue in play and means that
a party has the responsibility to ensure that there is sufficient evidence of
the existence or non-existence of a fact or an issue on the record to pass the
threshold for that particular fact or issue. Nu-Pharm, supra, per Stone
J.A., at page 33. [Emphasis added.]
[179]
According to the Supreme Court of Canada in R. v. Fontaine, 2004
SCC 27, [2004] 1 S.C.R. 702 at paragraph 11 [Fontaine], the “evidentiary
burden” is not a burden of proof. It is a legal question left for the judge to
determine whether “there is some evidence upon which a properly instructed jury
could reasonably decide the issue” (Fontaine, above, at para. 13). In
making such a determination, “the judge does not evaluate the quality, weight
or reliability of the evidence” (Fontaine, above, at para. 12).
[180]
Justice Wetston in Pharmacia Inc. v. Canada (Minister
of National Health and Welfare)(1995), 60 C.P.R. (3d) 328 at paragraph 28,
92 F.T.R. 253 (F.C.T.D.) [Pharmacia cited to C.P.R.], held that the presumption
of infringement is bolstered by the common law presumption. According to Pharmacia,
the common law presumption is (above, at para. 20):
[…] where the
subject-matter of an allegation lies particularly within the knowledge of one
of the parties, that party must prove it. In this instance, the applicants
submit that only the respondent knows the precise composition and process to be
used in making their product.
[181]
The Court of Appeal in Hoffmann-La Roche set out the test for
establishing the common law presumption: (a) the defendant asserted no facts to
support allegations of non-infringement; (b) the evidence of non-infringement
lay peculiarly within the knowledge of the defendant; and (c) the plaintiff had
no other available means to access such evidence (above, at para. 8).
[182]
Combining principles of the evidentiary burden, the common law presumption
and s. 39(2), I conclude that Apotex’s burden is to show that it used a
non-infringing process to create lovastatin, and that this process was
disclosed to Merck. At this time, the Court should not assess the quality,
weight or reliability of the evidence, but merely ask if there is sufficient
evidence to put the issue in play.
[183]
I find that Apotex, by its disclosure of the AFI-4 process (fermenting Coniothyrium
fuckelii to produce lovastatin), has met its evidentiary burden to rebut
the presumption of s. 39(2). Merck has accepted that the non-infringing AFI-4
process was used at the AFI plant in Winnipeg to produce lovastatin
(except for CR0157). Because Apotex has met its evidentiary burden, the
presumption of infringement has been rebutted. In the words of Wigmore, “… the
presumption disappears as a rule of law and the case is in the jury's hands
free from any rule” (Circle Film, above, at p. 606). The persuasive
burden of proof is back with Merck.
[184]
By way of summary, the burden within s. 39(2) of the Act is as
follows:
a)
Merck has the evidentiary burden to prove its substance is “new” in order
to engage s. 39(2);
b)
Apotex then has the evidentiary burden to prove a viable alternative
process existed to create lovastatin that does not infringe the '380 Patent –
if this is done, the presumption of law is lifted; and finally,
c)
Merck has the persuasive burden to prove infringement.
[185]
On the facts of this case, I am persuaded that the '380 Patent involves a
“new substance” thereby engaging s. 39(2). Apotex has established that a viable
alternative exists that does not infringe the '380 Patent. Accordingly, Merck
has the persuasive burden to satisfy me, on a balance of probabilities, that
Apotex’s sale of lovastatin, manufactured by Blue Treasure lovastatin or out of
batch CR0157, was made by a process that infringed the '380 Patent.
[186]
I turn now to consider whether Merck has met its burden. In my view they
have, with respect to some of the lovastatin manufactured by Blue Treasure and
lovastatin that originated from AFI batch CR0157.
C. Summary of Merck’s case on
infringement
[187]
As noted above, Merck claims infringement of the '380 Patent in three
different scenarios:
1.
infringement, between April 1997 and March 1998, through the manufacture
by Blue Treasure of quantities of infringing lovastatin that were shipped to
AFI, when Blue Treasure was allegedly “salting” the lovastatin shipments with
infringing Aspergillus terreus lovastatin;
2.
infringement from March 1998, when Blue Treasure was allegedly shipping
lovastatin manufactured with Aspergillus terreus; and
3.
infringement through the manufacture (during Phase 1 of production
described above), by AFI in Winnipeg, of quantities of lovastatin included in batch CR0157.
[188]
I will deal with each of these scenarios separately.
VII. Infringement
– the Circumstantial Case
A. Blue
Treasure “Salting”
[189]
Merck asserts that Blue Treasure was “salting” its earlier shipments of
lovastatin with infringing AFI-1 lovastatin. In simple terms, Merck submits
that Blue Treasure was “diluting” its AFI-4 lovastatin with AFI-1 lovastatin,
thereby infringing the '380 Patent with each and every shipment of lovastatin
to AFI.
[190]
On March 18, 2010, Merck received a copy of an e-mail, dated September 8,
1997 purportedly from Dr. Su to his “managers” at AFI. In the e-mail, Dr. Su
wrote that “before the switchover, [Blue Treasure] . . . produced 296.6 kg #1
product”. This 296.6 kg product is the basis of Merck’s salting argument. The
Defendants have not provided evidence as to how this quantity of AFI-1
lovastatin was sold or disposed of.
[191]
Does the failure of Blue Treasure to account for this quantity of AFI-1
lovastatin lead to a finding, on a balance of probabilities, that this
lovastatin ended up being shipped to Canada as a “salted” mixture
with non-infringing AFI-4?
[192]
Merck argues that, absent evidence of the whereabouts of the entire 296.6
kg, the reasonable inference is that all or part of that quantity of AFI-1
lovastatin was used to “salt” the AFI-4 lovastatin. By salting the Winnipeg shipments
with infringing lovastatin, Merck asserts that Blue Treasure was able to sell
the more cheaply-made lovastatin where payment was priced, on a kilogram basis,
on the more costly AFI‑4 lovastatin. Blue Treasure was also thereby able
to dispose of its aging inventory of infringing lovastatin that could not be
moved on the domestic Chinese market. According to Merck, if there were an
innocent explanation for Blue Treasure’s disposition of the infringing lovastatin,
AFI would have provided it; none has come.
[193]
In addition to the Defendants’ failure to provide sufficient information
on the disposal of 296.6 kg of AFI-1 lovastatin, Merck points to two other key
factors which, in their view, support the allegation of “salting”. The first
point is the evidence that demonstrates the difficulty that Blue Treasure was
having selling lovastatin into the Chinese or other foreign markets at a
reasonable profit. (This evidence is discussed at some length in the section of
these reasons dealing with “motivation”.)
[194]
Moreover, Merck refers to AFI’s concern that unusually low levels of RC-14
were found in the first two shipments of lovastatin from Blue Treasure. Merck
argues that AFI had been worried enough about the possibility of “adulteration”
that Dr. Su was sent to Blue Treasure to investigate and supervise the Blue
Treasure lovastatin productions. No determination was ever made by Dr. Su about
the reason for the unusually low levels of RC-14 in the shipments of AFI-4
lovastatin. Merck appears to argue that the RC-14 level could be explained if
the shipment to AFI had been, in fact, a mixture of AFI-1 and AFI-4 lovastatin.
[195]
In a letter dated August 11, 1997, Dr. Cox advised Mr. Zhou that:
I was very
disappointed to learn that 2 out of 3 batches of lovastatin made by the new AFI
process had failed our quality control, partly because your people had
unilaterally made changes to our process.
[196]
The most worrying problem with the lovastatin batches was the low levels
of a compound known as RC-14. In a letter from Mr. Alexander (Sandy) Fowler
(Finance Manager at AFI) to Mr. Zhou dated September 12, 1997, Mr. Fowler
described the “puzzle” as follows:
Our scientists are
very concerned in regard to the low levels of RC14 in certain of the batches
produced at Blue Treasure. At [AFI], we have never produced AFI-4 lovastatin with
such low RC14. In fact, in our experience, the “signature” of AFI-4 is a higher
level of RC14.
[197]
Low levels of RC-14 were characteristic of AFI-1. Thus, it is clear that
the real concern was whether the AFI-4 lovastatin was being produced using
AFI-1 or was being contaminated in some way by the Aspergillus terreus
strain. During his oral testimony, Dr. Cox was very direct in his testimony
about the lack of trust between AFI and Blue Treasure. The decision was made by
AFI to send Dr. Jerry Su to Blue Treasure to work with the management team (see
letter dated August 18, 1997 from Dr. Cox to Mr. Zhou).
[198]
Dr. Jerry Su was the Group Leader of Research & Development at AFI
from September 1996 to December 1998. Dr. Su arrived in China on August
28, 1997 and remained there until the end of October 1997. As set out in his
“Report on the work at Blue Treasure”, dated November 13, 1997, his
primary task was to ensure that Blue Treasure maintained all fermentation free
of contamination from the Aspergillus terreus strain. He investigated
the low levels of RC-14 but, even after his time in China and his
examination of a number of possible reasons, the cause of low RC-14 levels in
two batches was still a “puzzle”. Nevertheless, Dr. Su appeared to be satisfied
that the runs conducted while he was at Blue Treasure would meet the quality
control standards at AFI.
[199]
Together, Merck submits, all of this evidence is consistent with a finding
that, on a balance of probabilities, Blue Treasure was mixing infringing AFI-1
lovastatin with the AFI-4 lovastatin that was being shipped to AFI from Blue
Treasure.
[200]
In terms of the technical feasibility of salting, Merck relies on the
statement of Dr. Cox who testified that there is nothing difficult about
salting:
Q. You
understood, at the time, doctor, that Lovastatin made from one process could be
made and mixed with Lovastatin made using another process; you understood that
was a technical feasibility?
A. It's very
straightforward. You put them together and mix them.
[201]
I agree with Merck that the Defendants – in particular AFI – have
presented obstacles to uncovering relevant facts related to the amount of AFI-1
lovastatin actually produced and sold by Blue Treasure. Once the e-mail of Dr.
Su came to light on March 18, 2010, Merck sought and was granted, on consent,
an opportunity for further discovery of Mr. Fowler. Mr. Fowler has been
the Finance and Administration Manager at AFI since 1996. Questions related to
this lovastatin were put to Mr. Fowler and taken under advisement by counsel
for AFI. AFI refused to provide confirmation of the 296.6 kg quantity
(Undertaking #2319). In Undertaking #2320, AFI was also asked, in respect of
the 296.6 kg product:
To advise full
particulars, when it was made, what happened to it, who it was sold to, for how
much and financial benefits to AFI and Apotex.
[202]
All of this information was refused. Although Apotex refers to some
evidence on the record that directionally supports legitimate sales of AFI-1
lovastatin, the information is far from complete or clear.
[203]
In spite of my serious concerns about the unwillingness or inability of
the Defendants to provide evidence concerning the alleged 296.6 kg product, I
have problems with Merck’s argument on this point.
[204]
My first concern is that I have little evidence that Blue Treasure
produced 296.6 kg of AFI-1 lovastatin. In Dr. Su’s e-mail, there is the
reference that, “before the switchover, [Blue Treasure] . . . produced 296.6 kg
#1 product”. There is no indication in the e-mail of where this number came from.
Contrary to the assertion by Merck, the e-mail does not “prove” the existence
of 296.6 kg of infringing lovastatin.
[205]
The second problem is that I have no evidence, beyond the cryptic
statement of Dr. Cox, as to how salting could be carried out. No expert spoke
to the practice. As noted by counsel for Apotex in final argument:
. . . there was no
evidence led as to how you put together AFI‑1 material with AFI‑4
material, whether the appearance and composition of technical grade lovastatin,
colour wise, physical characteristics, crystallinity, whether that is
comparable.
[206]
In addition, I have no confirmation from anyone that low levels of RC-14
could be explained by salting. That question could have been posed to any
number of witnesses by Merck and was not. The closest discussion on the record
occurred during the cross-examination of Mr. Fowler. Mr. Fowler was
questioned at length about why Dr. Su was sent to China.
Specifically, the concern that Blue Treasure had switched the organisms was put
to Mr. Fowler. His response included a vague reference to the possibility of
“contamination or mixing of product”:
The concern from my
perspective was that there not be any mix‑up or errors resulting in a
contamination or mixing of product, that sort of thing. Certainly, in
everybody's mind there was a possibility of some mix‑up, and that's what
we wanted to get to the bottom of.
This statement is certainly not
sufficient for me to conclude that AFI believed that “salting” was going on at
Blue Treasure.
[207]
I conclude that it is possible that Blue Treasure used some of the alleged
296.6 kg of AFI-1 lovastatin to “salt” the AFI-4 lovastatin being shipped to
AFI. However, on the evidence before me, I have insufficient evidence about how
that could have been done from a technical perspective. Nor do I have any
evidence that AFI believed that “salting” might have been the reason for the
low levels of RC‑14.
[208]
In short, Merck has failed to persuade me, on a balance of probabilities,
that any quantities of lovastatin were salted with infringing lovastatin.
B. Infringement by Blue
Treasure from March 1998
[209]
As I understand Merck’s argument on infringement after March 1998, the key
reasons why I should find that the lovastatin manufactured by Blue Treasure and
sold to AFI for sale in Canada, was produced using a process that infringed the
'380 Patent are as follows:
·
The Batch Records (referred to below) produced by Blue Treasure to
demonstrate the use of the non-infringing AFI-4 process for producing
lovastatin are not genuine.
·
The Defendants failed to provide evidence that Blue Treasure acquired
sufficient quantities of a compound known as P2000 with which to carry out
fermentations with the Coniothyrium fuckelii micro-organism.
·
The reduction in the duration of fermentation, beginning in March 1998, is
unexplained except by the use of the infringing process.
·
The increase in the quantities of titres, beginning in March 1998, is
unexplained except by the use of the infringing process.
·
Blue Treasure had the motivation, the means and the opportunity to produce
lovastatin with the less expensive, more efficient AFI-1 infringing process.
·
The conduct of Blue Treasure before and during this trial is consistent
with infringement.
[210]
In addition to the above – much of which consists of circumstantial
evidence – Merck asserts that it has direct evidence of infringement through the
DNA evidence put forward by Dr. Julian Davies.
(1) Batch
Records
[211]
Standard pharmaceutical industry practice requires that detailed and exact
records be kept of all steps in the production of pharmaceutical products. In
this trial, the Defendants put forward 364 documents as true photocopies of the
batch records for 364 fermentation batches of lovastatin made at Blue Treasure
(the Batch Records). There is no doubt that, if the Batch Records (all of which
are contained in Exhibit 149) can be believed, they are direct evidence that
Blue Treasure was using the non-infringing AFI-4 process and not a process
using Aspergillus terreus. However, the issue is whether I can believe
that the Batch Records are reliable, or even truthful, in some important
aspects. Quite simply, I cannot.
[212]
The first problem with the Batch Records is that they are not the original
working records from the batches. The original Batch Records, together with
every single related document and working paper, were destroyed in 2003. All
that is before this Court are photocopies of what is alleged to be the original
batch records.
[213]
The Batch Records were introduced into evidence by Mrs. Quifen Hu. Mrs. Hu
has been the Manager of the Bacterial Culture Department at Blue Treasure since
1995. From 1991 to 1994, she was the Manager of the Intermediates Department at
New North River in China. Since 1995, Mrs. Hu’s responsibilities have mainly
related to maintaining the seed bank at Blue Treasure and initializing the
fermentation process for the production of lovastatin.
[214]
Each of the 364 Batch Records is substantially the same in format; each
appears to be a pre‑printed form with data and other entries written by
hand. They reflect fermentations that began on May 27, 1997, with batch no.
CF-403-97001, and ran until the fermentation of batch no. CF‑410‑99166,
which began on September 29, 1999.
[215]
Parts 1.1 and 1.2 are entitled “Production Record of Lovastatin
Fermentation”. The list of ingredients of the medium is printed on the form and
the quantities used are hand written in the appropriate space. Information on
the pH adjustment and the steps for inoculation of the seed flasks are
documented. The final step described is the culture of a second seed flask.
[216]
Parts 2 to 14 of each batch record set out the scale-up of fermentation from
the final step of Part 1.2 to Part 13, which is entitled the “Production
Fermenter”, where the final product is fermented. Along the way, increasingly
larger fermenters are prepared and then inoculated with the seed media from the
previous stage.
[217]
The only place that the initiating or seed micro-organism is described is in
Parts 1.1 and 1.2. The batch no. is recorded on the cover sheet of each batch record
and on most pages, and always begins with the letters “CF”. Presumably, this
means that the particular run is being carried out with Coniothyrium
fuckelii. In addition, each of the 364 Batch Records is pre-printed with an
entry that identifies the production strain as “Coniothyrium fuckelii
AFI-4 85-42”. For Parts 2 to 14, no strain is identified.
[218]
Mrs. Hu was responsible for overseeing the process to the end of Part 1.2.
After that, the production steps were within the responsibility of the Fermentation
Production Department, where Mr. Dingjun Luo (whose testimony is referred to
later in these Reasons) worked.
[219]
The first point to make about the Batch Records is that they do not
qualify as business records that can be accepted for the truth of their
contents as an exception to the hearsay rule of evidence.
[220]
Since Ares v. Venner, [1970] S.C.R. 608 at p. 363, 12 C.R.N.S. 349,
the common law in Canada has recognized that certain records:
. . . made
contemporaneously by someone having personal knowledge of the matters then
being recorded and under a duty to make the entry or record should be received
in evidence as prima facie proof of the facts stated therein.
[221]
The common law has been codified in s. 30(1) of the Canada Evidence Act,
R.S.C. 1985, c. C-5:
|
Where oral evidence in respect of a matter would be
admissible in a legal proceeding, a record made in the usual and ordinary
course of business that contains information in respect of that matter is
admissible in evidence under this section in the legal proceeding on
production of the record.
|
Lorsqu’une preuve orale concernant une chose
serait admissible dans une procédure judiciaire, une pièce établie dans le
cours ordinaire des affaires et qui contient des renseignements sur cette
chose est, en vertu du présent article, admissible en preuve dans la
procédure judiciaire sur production de la pièce
|
[222]
The evidentiary record on the Batch Records is, to put it bluntly, a mess.
One thing that I can say with certainty is that the Batch Records reflected in
Exhibit 149 were not made contemporaneously with the fermentations to which
they ostensibly relate. I need only compare these Batch Records to those
produced as part of the production runs at AFI. The AFI records contain
numerous corrections and deletions. From the writing styles, it is clear that a
number of persons completed the documents at various stages of each
fermentation. In contrast, the Batch Records do not contain a single strike-out
or correction; they, as described by counsel for Merck, are “pristine”. This
would be unheard of in any production facility.
[223]
As I understand it, AFI acknowledged, late in the trial, that the staff at
Blue Treasure completed unilingual Chinese work sheets on the plant floor and
that, subsequently, the data collected was transposed into bilingual batch
records. Other evidence, provided through an Undertaking, is that Mr. Luo
confirmed that unilingual sheets “did exist and that information from those
sheets . . . were transferred into the bilingual batch record”.
[224]
If these are not business records, the question becomes: how much, if any,
weight should be given to the Batch Records?
[225]
Three witnesses provided testimony on the Batch Records – Mr. Dingjun Luo,
Mrs. Hu and Dr. Jerry Su.
[226]
The best evidence of how the Batch Records were created, and by whom, was
provided by Dr. Jerry Su. As noted earlier, from September 1996 to December
1998, Dr. Su was the Group Leader of Research & Development at AFI. He has
not worked for AFI since that time. Dr. Su was quite firm in his recollection,
having spent long hours at the Blue Treasure plant observing and taking notes
as to the daily operations. The following flows from Dr. Su’s testimony:
·
Blue Treasure maintained two sets of batch records. There were no
“unilingual worksheets” for part of the process, only the original Chinese
language batch records, which Dr. Su called the “first set” of batch records.
These were used on the plant floor and entries were made therein by operators.
·
The first set of batch records was collected and the data therein were
copied out, by hand, into a “second set” of bilingual batch records.
·
Dr. Su sat next to, and witnessed, the person copying the batch records,
but did not see whether the information was accurately being copied.
·
The person who copied the data into the second set of batch records (apparently
the originals of the ones before the Court) was Mr. Luo, Manager of the Production
and Technology Department.
·
Dr. Su did not recall ever seeing Mrs. Hu signing any of the Batch
Records.
[227]
I have no reason to disbelieve Dr. Su and, where there is conflicting
testimony from Mrs. Hu or Mr. Luo, I prefer the testimony of Dr. Su. At present,
Dr. Su has no connection with AFI or Blue Treasure and no motive to fabricate
his evidence. His explanation of the transfer of data from the original working
documents is logical and consistent with the form of the Batch Records that
make up Exhibit 149.
[228]
Mr. Luo’s testimony was presented by AFI in an attempt to validate the
Batch Records as reliable documents. Mr. Luo is currently Deputy General
Manager with Blue Treasure. He first joined Blue Treasure, as a technician, in
1995. In 1996, he described his position at Blue Treasure as Head of Production
and Technology.
[229]
Mr. Luo was a very difficult witness. AFI’s counsel admitted, in oral
final argument, that there were problems with Mr. Luo’s testimony. As the
evidentiary record with respect to the Chinese Journal Articles demonstrates
(see para. 322 of these reasons), Mr. Luo was prepared to fabricate evidence
when it served his purpose. His testimony was replete with poor memory of
matters that ought to have been known to him. In spite of his senior position
at Blue Treasure, he claimed to be unaware of any matter that did not fall
directly under his supervision. His evidence on the subject of the Batch
Records was particularly problematic.
[230]
When cross-examined on the question of the second set of Batch Records,
Mr. Luo was unequivocal:
Q. Thank you,
sir. Did you tell the lawyers that there was a unilingual Chinese worksheet
from which data were copied into a second set of batch records?
A. There is no
such first set and second set of records.
Q. Did you tell
the lawyers that a unilingual worksheet, containing the parameters mentioned in
a letter from February, were prepared and that those data were copied into the
second set of batch records which have been produced in the litigation? Did
you tell the lawyers that?
A. No.
[231]
The problem is that Mr. Luo provided a different and conflicting
explanation about the Batch Records, recorded as a response to Undertaking #
6585:
Mr. Luo advised that
it was the operators that transferred information from the unilingual
worksheets into the bilingual English/Chinese batch record. He would then check
the bilingual batch record to ensure that it was accurate and that no
information was missing.
[232]
As demonstrated by these examples and other portions of the record, the evidence
of Mr. Luo with respect to the Batch Records lacks credibility.
[233]
Another witness who spoke about the Batch Records was Mrs. Hu. Apotex submits
that I should accept Mrs. Hu as a reliable witness. I find that difficult to
do.
[234]
Mrs. Hu was an evasive and difficult witness. Her testimony on the Batch
Records was no less confusing than that of Mr. Luo. Repeatedly, she refused to
acknowledge matters that should have been within her knowledge. She was lead
extensively through her direct testimony. In general, her testimony in chief
was a reading of the Batch Records. Mrs. Hu’s recollection of making lovastatin
with Coniothyrium fuckelii from May 1997 to October 1999 and her testimony
with respect to the Batch Records were not consistent and did not stand up on cross-examination.
[235]
When taken to the cover page of the batch record for the first run of
AFI-4, Mrs. Hu testified that she signed the original of this batch record on
May 26, 1997, the date reflected on the record. Similarly, she testified that
she signed each and every one of the 364 Batch Records on the date that the
various operations were performed. She vehemently and – in my view –
illogically clung to her testimony that she herself signed the record on each
and every day. Mrs. Hu also testified that there were no unilingual Chinese
records and that she had never seen anyone writing numbers on a separate work
sheet for copying into the batch records later. This, of course, is not what
the Court was told by Dr. Su. Moreover, her testimony that there were no
unilingual or working records is inconsistent with the “pristine” appearance of
the Batch Records in Exhibit 149.
[236]
Mrs. Hu may well have signed at the appropriate spots in Parts 1.1 and 1.2
of the reconstructed Batch Records. She may well have believed that Parts 1.1
and 1.2 of the Batch Records accurately reflected what she had done at that
stage of the fermentation runs. However, I find, as a matter of fact, that she
did not do so contemporaneously with the fermentation runs. Rather, she signed
the pages after Mr. Luo had completed the forms. It is difficult to say whether
Mrs. Hu, whose testimony demonstrated no understanding of the English language,
was even able to read what was printed and written on the forms or took the
time to compare the information on the original operational records to the
headings and data that appeared in Parts 1.1 and 1.2 of the Batch Records. The
original records could have helped corroborate her testimony; sadly, they were
destroyed. Most significantly, Mrs. Hu’s testimony does not reliably establish
which runs, if any, were made using AFI-4, in spite of the use of the
identifier “Coniothyrium fuckelii AFI-4 85-42” on each of the Batch
Records.
[237]
In sum, I am not persuaded that the Batch Records contain the original data
from the 364 fermentations allegedly carried out by Blue Treasure; rather, they
were transcribed from the originals, most likely by Mr. Luo.
[238]
Another area of concern regarding the Batch Records was the destruction of
the original documents in January 2003. The fact that the original batch
records were destroyed six years into this litigation by Blue Treasure, a joint
venture in which the Defendants are partners, raises a serious concern. Were
they destroyed, as asserted by Merck, to hide evidence of the actual production
methods?
[239]
Mrs. Hu’s testimony on why the original batch records were destroyed was
inconsistent and illogical. According to Mrs. Hu, she was the person who
authorized the destruction of the originals. She testified that she did so
pursuant to a Chinese GMP policy (apparently entitled “Guidance for Industry –
Q7A Good Manufacturing Practice Guidance for Active Pharmaceutical
Ingredients”). However, when questioned about the policy, it was clear that she
did not have even the most basic understanding of the policy. Mrs. Hu was
unable to recall clearly any of the details related to the destruction. She
could not remember clearly the last time she signed a request for destruction.
Finally, the copy of the GMP Policy presented to the Court was apparently
brought into effect in April 2003 – three months after the alleged
destruction. I am left without any reasonable explanation as to why the
original records were destroyed.
[240]
I find that the original batch records were destroyed for reasons that cannot
be determined, thereby leaving me unable to confirm any of the information
contained in the Batch Records.
[241]
Moreover, any further confirmation of how and when the Batch Records were
actually prepared is impossible. Mr. Brian Lindblom, recognized by the Court as
an expert in forensic document examination, provided helpful opinions regarding
this issue. Mr. Lindblom examined the Batch Records and, in his Expert Report,
provided the following list of relevant tests that he would have performed on
the original documents had they been available to him (Lindblom Expert Report,
Exhibit 66, para. 19):
(a) whether
the paper used in the original batch record was in fact available at the time
the documents were allegedly made;
(b) whether
all of the inks used in the original batch record were in fact available at the
time the documents were allegedly made;
(c)
whether the ink has been on the document for the amount of time indicated
by the date of the batch record;
(d)
whether a single ink had been used for a batch record supposedly completed
by numerous people over many days (in which one would expect to find more that
one ink type);
(e)
whether and perhaps when alterations had been made to documents which
might suggest some type of fabrication;
(f)
whether the documents were in fact completed in the chronological fashion
suggested by the time line set out in the batch records (this can be determined
through examination of indentations); and
(g)
whether more than one person completed the handwriting on the batch record
(as one would expect for a process which occurred over many days).
[242]
I conclude that the Batch Records are not reliable or trustworthy evidence
that the fermentation runs that took place at Blue Treasure after March1998 were
using the AFI-4 process to produce lovastatin. The pre-printed forms could have
easily referred to Coniothyrium fuckelii as the production strain, even if
a strain of Aspergillus terreus was used. The data could readily have
been changed to cover up the use of the AFI-1 infringing process. Indeed, the
lack of credibility of the two Blue Treasure witnesses leads me to conclude
that it is more likely than not that the Batch Records were fabricated, at
least with respect to any information that could identify the strain of
micro-organism used.
[243]
The Batch Records and the testimony of Mr. Luo and Mrs. Hu are critical
underpinnings of the Defendants’ defence to the allegation of infringement. In
my view, the Batch Records contain data about the fermentations that are not
consistent with that defence.
[244]
Apotex asserts that Merck’s argument that the Batch Records should not be
relied on is inconsistent with Merck’s use of the data to highlight the problems
with the titres, the use of P2000 and the reduction in fermentation duration. I
do not find the position of Merck, when explained in oral argument, to be
inconsistent. I believe that what Merck is saying is that the Batch Records
should be completely rejected. In the absence of reliable records on the
fermentations, Merck asserts that the Defendants have no defence to the
allegation of infringement. However and alternatively, Merck argues, if the
Batch Records are to be believed, the data demonstrates that Blue Treasure must
have been using the infringing AFI-1 process after March 1998.
[245]
On a final note, the Defendants submit that I should exercise caution in
assessing the credibility of Mrs. Hu and Mr. Luo. In particular, they point to
the existence of cultural differences that could account for the behaviour or
demeanour of these witnesses.
[246]
This concern about applying “western” standards to an assessment of
credibility was considered, in a criminal context, by the Ontario Court of
Appeal in the case of R. v. E.(T)., 2007 ONCA 891, [2007] O.J. No. 4952
(QL) [R. v. E.(T).]. In that case, the trial judge stated that he did
not believe the accused, relying heavily on the demeanour of the accused during
his testimony. The Court of Appeal at paragraph 5, in finding that the judge
erred, stated as follows:
The appellant contends
that the trial judge's references to the appellant's apparent passivity and to
his failure to make eye contact with the other witnesses at the trial
constitute erroneous use of demeanour evidence. We agree. As the trial judge
himself observed, accused persons can react differently in a stressful criminal
trial. Without explaining why, and without acknowledging the effect of cultural
background on demeanour (the appellant was born and raised in Sudan), the
trial judge equated passivity and an absence of eye contact with witnesses with
rejection of the appellant's credibility and, ultimately, his testimonial
denial of committing the offence. This equation is, in our view, a misconceived
and improper linkage.
[247]
In the criminal context, there is a significant difference in burden and
standard of proof. The Crown, in R. v. E.(T.), bore the burden of
proving all of the elements of the alleged offence beyond a reasonable doubt.
It was open to the Crown to put forward expert evidence on cultural background;
had it done so, the outcome in the Court of Appeal’s decision might have
differed. Further, it appears, from reading the judgment of the Court of
Appeal, that the trial judge placed much of his reliance on the demeanour of
the accused; as noted by the Court of Appeal, the trial judge “equated” demeanour
with rejection of the accused’s credibility.
[248]
In the case before me, Mrs. Hu and Mr. Luo are the Defendants’ witnesses.
They were put forward expressly to respond to the claim by Merck that Blue
Treasure was, at least in part, using the infringing AFI-1 process. During
their appearances, I was never alerted to the fact that any cultural background
issues would alter how they presented themselves. Nor did the Defendants put
forward any expert testimony on what these alleged “cultural differences” could
be. In fact, beyond a general caution, the Defendants make no attempt to
describe what particular attributes would affect the testimony or to describe
what specific aspects of their demeanour reflect a cultural difference. In the
circumstances, I can see no reason why I should not rely on demeanour (within
reason and not solely) in assessing their credibility.
[249]
Having said this, however, I am aware that both Mrs. Hu and Mr. Luo
testified through an interpreter. Even with competent translation (which we had
in this trial), I can appreciate that there may be times where the witnesses
could have become frustrated with the inability to testify directly. This
frustration was readily apparent with Mrs. Hu – less so with Mr. Luo. In those
situations, the answers may not have been as clear as they could have been. I
have taken that into account.
(2) P2000
[250]
The evidence before me is that the compound known as Polyglycol P2000 (P2000)
was an essential ingredient in the AFI-4 process. As I understand it, foaming created
by the fermentation process negatively affects the production levels. When P2000
was added to the fermenters it acted as an anti-foaming agent and increased the
titres dramatically. Without P2000, it is fair to say that the production of
lovastatin using Coniothyrium fuckelii would not be commercially viable.
[251]
Dr. Connors, an expert witness who provided the Court with his
understanding of the history of the AFI-4 process at AFI, spoke in superlatives
about the decision to add P2000 to the process. According to Dr. Connors, the
addition of P2000 to the fermentation medium in March 1995 was a key breakthrough
in the development of the AFI-4 process at AFI:
This is March of
1995. This is really the big break in the project. I will take you to, under
tab 63, AFI production 4-39, the very first one. This is when you are glad to
be a scientist. This is when you are glad that you chose science. This,
again, is culture 115 57, and this is the first data that suggests that adding
P2000 in larger amounts than what you would normally add it gives you a
dramatic increase in the production of Lovastatin with this culture.
You can see through
the top row of data that day six, day seven, day eight titers, 300 mgs per
litre. If you increase the P2000 up to two percent or 20 mls per litre, by day
eight you have almost 1.6 grams per litre. Spectacular. I mean, this is
incredible.
You can see, also,
more importantly too, or just as important, that this is a dose dependant
manner. As you increase the concentration of P2000 incrementally, you see a
concomitant increase in the titer of Lovastatin as well. So this was a very
good result. This was something that really broke the project open for them.
[Emphasis added.]
[252]
The increased amount of P2000 needed for the AFI-4 process as compared to
the AFI-1 process is dramatic – in the order of 10 to 20 times more is required
for the AFI-4 process. Mr. Scott Primrose is a senior research scientist at
AFI. In 1993, he was working in the microbiology laboratory of AFI, where much
of his work focussed on the development of non-infringing lovastatin – that is,
the AFI-4 process. He confirmed that the AFI-1 process required about 1/20th
of the amount of P2000. Dr. Sailer described the amount of P2000 used for the
AFI-4 process as “very unusual”:
Typically amount of
anti‑foam used for fermentation it's like point two percent and in this
case was two percent, 10 times more. [Emphasis added.]
[253]
Merck submits that Blue Treasure did not have sufficient quantities of
P2000 to complete 364 runs of the AFI-4 process. Merck calculated that Blue
Treasure would have needed 73,338 kg of P2000 to run 364 fermentation batches.
This calculation was not disputed by the Defendants. Blue Treasure also needed
P2000 for other projects, including for the production of compactin and AFI-1
lovastatin.
[254]
How much P2000 did Blue Treasure have? According to the evidence, between
May 1, 1997 and June 9, 1998, AFI made 10 shipments of P2000 to Blue Treasure
for a total of 27,784.80 kg. This would have been far short of the 73,338
kg needed to run the AFI-4 process, but certainly sufficient to supply the
AFI-1 process. Approximately 5 to 10% of the amount of P2000 is needed for the
AFI-1 process. How did Blue Treasure obtain the remaining P2000 that it needed?
There really was no answer to that question.
[255]
Dr. Connors speculated as follows:
The balance of the
P2000 could have come from someplace else. Might have been the amount they
started with. Perhaps this was the amount of raw material that Winnipeg provided,
and P2000 was sourced through some local vendor. P2000 is not an unusual
chemical. It's a commodity, and could very well be available in China. I don't
know for sure.
[256]
The Defendants assert that, after July 1998, Blue Treasure began to source
its own raw materials, including P2000. Mr. Fowler, speaking to this issue
during his testimony, stated that “after the fifth order [for P2000] Blue
Treasure was able to purchase its own raw materials”. However, Mr. Fowler
provided nothing beyond speculation that Blue Treasure was accessing P2000 from
other markets; he had no personal knowledge of how that might have happened or
if, indeed, it did.
[257]
Beyond the speculation of Dr. Connors and Mr. Fowler, I have nothing that
speaks to Blue Treasure’s acquisition of the required quantities of P2000. I
have seen no invoices or shipment statements to back up the Defendants’
assertions in this regard.
[258]
The missing P2000 is extremely important. Without sufficient quantities of
P2000, Blue Treasure could not have produced AFI-4 lovastatin throughout the
entire 364 runs. There are obviously people associated with Blue Treasure who
could have provided evidence of additional purchases of P2000, if such
purchases had taken place. For example, Mr. Zhou was the Plant Manager at the
relevant time. Of those who might have been able to assist the Court, only Mr.
Luo was called as a witness and he was not asked about P2000. In the
circumstances, I will presume that the evidence that could have been provided
by the witnesses from Blue Treasure, who were not called to testify, would have
adversely affect the Defendants’ case (see Levesque v. Comeau, [1970]
S.C.R. 1010, 16 D.L.R. (3d) 425). In other words, I will assume that Blue
Treasure has no further evidence that it ever purchased enough P2000 to carry
out the 364 fermentations using the AFI-4 process.
[259]
In the absence of evidence to the contrary, it is not unreasonable to
believe that Blue Treasure was not using the AFI-4 process as the Defendants
claim. On the other hand, I have evidence that the Blue Treasure had sufficient
P2000 - 27,784.80 kg – to carry out the fermentations using the AFI-1
infringing process. I have persuasive evidence – albeit circumstantial – that
supports a conclusion that Blue Treasure was using the infringing AFI-1 process
to produce lovastatin.
(3) Fermentation
Duration
[260]
One of Merck’s arguments focuses on the notion of “fermentation duration”.
As discussed in the background section of these reasons, the production of
lovastatin requires fermentation over a period of time. It is self-evident to
say that a manufacturer will try to produce the largest volume of final product
over the least amount of time. For example, in general, producing 100 units of
material over 10 days is economically more efficient than producing the same
100 units over 12 days.
[261]
Production of lovastatin is reported in “titres” (also “titers”) – that
is, the concentration of a solution as determined by titration, usually
measured in units of mg/L or g/L.
[262]
Dr. Mila Sailer explained “fermentation duration” as the interaction of a
number of factors. Dr. Sailer was a fact witness presented to the Court by AFI.
Dr. Sailer worked as a natural product chemist with AFI, and was involved in
the production of lovastatin in October 1996 when AFI started its first
commercial production of lovastatin from Coniothyrium fuckelii. Dr.
Sailer explained that production of a product such as lovastatin requires
consideration of a number of factors:
There is a number of
factor which is to be considered when you, when you want to optimize the
production of the facility, fermentation facility. So, for instance, you have
to look on the production curve doing the fermentation run, you have to look on
a capacity of the seed train which is used for inoculation of production
vessels. You have to look on the capacity of the downstream equipment. You
have to look on, for instance, on impurity profile during the fermentation run,
which can change. You have to look on cost of the media.
[263]
Under the notion of the “production curve”, Dr. Sailer spoke about
fermentation durations and confirmed that often there is a trade-off between
titres and fermentation times. Sometimes, he said “it’s not the best to . . .
go for maximum titres”; rather, “sometimes it’s much better to shorten the time
and do a higher number of fermentation[s]”. While Dr. Sailer was not presented
as an expert, his testimony codifies common sense when it comes to a general
view of the factors affecting production.
[264]
The experience of AFI, in its Winnipeg facilities, during
the period between August 1996 and August 1997, was that 76 batches of AFI-4
were run with an average fermentation duration of 11 days. In other words, the optimization
spoken of by Dr. Sailer for AFI-4 was reached by an 11‑day fermentation.
[265]
As we know, Blue Treasure began its runs of AFI-4 in June 1997. About 70
runs were made at Blue Treasure between June 1997 and October 1997. If the
Batch Records are to be believed, although there is significant fluctuation,
the average fermentation duration during that period was about 11 days. This
was an expected result, given the experience of AFI with the AFI-4 process in Winnipeg.
[266]
Blue Treasure shut down its production of lovastatin from October 1997 to March
1998. Other than 17 runs that took place between December 1997 and January
1998, there was no production of lovastatin in this period. Upon resumption of
production in March 1998, the fermentation duration was immediately lower and
eventually stabilized at nine days.
[267]
These data were depicted by a graph, in Exhibit 83, prepared by Merck and
presented to, and discussed by, a number of witnesses. What could account for
the change in the fermentation duration?
[268]
Merck’s explanation for the reduction in the fermentation duration after
the plant shutdown is that Blue Treasure began to use the infringing AFI-1
process, instead of the AFI-4 process, to produce lovastatin.
[269]
The graph was first presented to Dr. Cox during cross-examination. Dr.
Cox, whose testimony was very credible and trustworthy, agreed with counsel for
Merck that Exhibit 83 showed that the AFI’s 11-day fermentation duration for
AFI-4 was roughly consistent with the fermentation duration experienced by Blue
Treasure up until the plant shutdown. He also agreed that the graph in Exhibit
83 depicted an average nine-day fermentation duration after the resumption of
production at Blue Treasure in March 1998. Finally, and most importantly, Dr. Cox
confirmed that the AFI-1 process transferred to Blue Treasure could be run in
nine days.
[270]
In my view, the decrease in fermentation duration after the Blue Treasure
plant shutdown is more likely to have occurred with the production of AFI-1
lovastatin than with the production of AFI-4 lovastatin. Thus, even if the
Batch Records are accepted as reliable evidence of the production runs, the
data with respect to fermentation duration are consistent with the use of the
AFI-1 process and not the non-infringing AFI-4 process.
(4) Increased
Titres
[271]
Fermentation duration is not the only variable to be considered in the
economics of producing lovastatin. The amount of production or titres from each
fermentation batch is of equal significance. Related to this issue is the role
of Amicase in the fermentation medium. The AFI-4 process transferred to Blue
Treasure required the use of a compound in the medium called Amicase. As
I understand it, Amicase is an enzyme that is used as a catalyst in a chemical
reaction. Dr. Connors testified that, “If you take out Amicase,
then you get less lovastatin.”
[272]
Amicase was an expensive ingredient and the evidence suggests that Blue
Treasure was taking steps to try to avoid its use. In a memorandum dated
October 23, 1997 to Dr. Xinfa Xiao and Dr. David He, both employees of AFI, Mr.
Huigen Xu, of Blue Treasure, explained that, when Blue Treasure tested
fermentation without Amicase, they “did not find any effect on the product
quality”. However, the same memorandum reports a reduction in titres of 9.75% at
the 12,800 L fermentation stage. Mr. Fowler confirmed that, although a loss of
productivity of 9.75% was the result of removing Amicase from the medium, Blue
Treasure could achieve a cost savings of about $224.00 US per kilogram by its
elimination.
[273]
From about January 1998, Blue Treasure produced lovastatin without using
Amicase. One would logically expect that thereafter there would be a
reduction in titres as reflected in the memorandum of October 23, 1997. That
does not appear to have been the case.
[274]
In Exhibit 103, Merck’s counsel attempted to pull together all of the
information on titres. This exhibit was prepared from Exhibit 149 (a complete
set of the Blue Treasure Batch Records), and presented in graphical format in
Exhibit 83. During three periods, the average titres recorded by Blue Treasure
showed the following:
|
Period
|
Average titres (g/ml)
|
|
June 7, 1997 to October 27, 1997
|
2.3 (n = 53 runs)
|
|
December 8, 1997 to January 11, 1998
|
2.0 (n = 17 runs)
|
|
March 7, 1998 to October 7, 1999
|
2.2 (n = 292 runs)
|
[275]
The chart demonstrates that there was a drop in titres of about 13% for
the period December 8, 1997 to January 11, 1998. However, the titres
increased, after the Blue Treasure plant shutdown, to a level approximately equal
to the runs made with Amicase. The obvious question is this: in the post-March 1998
period, how could Blue Treasure have obtained production levels without Amicase
consistent with those with Amicase?
[276]
Merck submits that the only reasonable explanation is that after March 7,
1998, Blue Treasure was using the AFI-1 process and not the AFI-4 process.
[277]
In final oral argument, Apotex refers to a sub-set of the Batch Records to
argue that there is no evidence that the omission of Amicase resulted in a reduction
in titres. For support of this statement, they refer to Table A of Exhibit 156,
listing “All L4-39-581 Fermentations Pre-March 1998.” Apotex points out that
the average titres of the three runs with Amicase are actually lower than in
the 12 runs listed in Table A that were run without Amicase:
The point is, using
the same time point from the data, when you remove Amicase at nine days, you
have a significantly better titre performance. So the argument, the argument
that you would expect titres to go down, because they did remove Amicase in the
later runs in March and following, is not borne out by this document which my
friend put forward as Exhibit 156. It shows exactly the opposite.
[278]
The problem with Apotex’s submissions on this point is that the average of
titres with Amicase highlighted by Apotex was based on an average of only three
runs (September 24, 1997, September 26, 1997 and October 17, 1997). Apotex
then compares this extremely small sample size against another small sample of
12 runs. In my view, the averages used by Apotex are based on a sub-set of the
Batch Records from which relative conclusions as to titres for the entire Batch
Records cannot be made.
[279]
I prefer the information set out in Exhibit 103, which demonstrates an
increase in titres of 13%. I accept that this is inconsistent with the omission
of Amicase.
[280]
Even discounting the effect of the omission of Amicase, there are
inexplicable changes in production levels after the plant shutdown. This change
in titres can be viewed in the context of L4‑39-581, one particular
strain of Coniothyrium fuckelii allegedly used both before and after the
plant shutdown. From the Blue Treasure Batch Records, we can track the runs
that Blue Treasure claims it used for this particular strain. Merck counsel
prepared Exhibit 156 from the Batch Records. Table A of Exhibit 156 is a
summary of the Batch Records for all runs using the L4-39-581 strain prior to
the plant shutdown. Between September 24, 1997 and January 11, 1998, 15
fermentations were carried out. Although these batches ran for a duration of 11
days, we can calculate the hypothetical titres at nine days from the records. If
these batches had been run for nine days, the average (arithmetic mean) titre
would have been 1719 mg/L. Table B of Exhibit 156 consists of information from
the Blue Treasure Batch Records for the alleged runs using the L4-39-581 strain
after the plant shutdown. From March 7, 1998 to April 23, 1998, 19 runs were performed
with an average fermentation duration of 8.3 days and an average titre of 2109 mg/L.
This shows an increase of over 20% in titres.
[281]
This increase is dramatic. Allegedly using the same strain of Coniothyrium
fuckelii, L4‑39‑581, Blue Treasure managed to increase the
productivity of the fermentation by over 20%. What explanation is there for the
increase? Merck submits that the only reasonable explanation is that, for the runs
after the plant shut down, Blue Treasure was not using a strain of Coniothyrium
fuckelii, but was using a strain of Aspergillus terreus.
[282]
In support of its belief, Merck referred to the testimony of Dr. Connors
who was presented with scatter graphs representing the variations in titres
among the Blue Treasure runs. Dr. Connors agreed that “one possible
explanation” for a variation in titres before and after the plan shut down, could
be the micro-organism or fungus used, when the medium and the process are kept
constant. Dr. Connors did not specify what other explanations there could
have been other than to say that something had to be different:
What did you do in the
middle that you didn’t do on either side? I wouldn’t necessarily say the
culture has to be different, but something is different, obviously.
Dr. Connors provided no further
assistance to the Court on what could explain the differences.
[283]
The data replicated in some of the exhibits produced by Merck’s counsel
from the Batch Records show that the variability in titres between runs diminished
over time. I can accept, as a matter of common sense, the comments of Dr. Cox
who explained that, “with the passage of time and the practice and the
reproducibility and the adherence to the [Standard Operating Practices] and the
ironing out of the problems”, there would be more consistency in titres.
However, this does not account for the dramatic and immediate change in titres
in evidence before me.
[284]
The only witness presented by the Defendants who could speak to the dramatic
change in titres was Mr. Luo. He was on the plant floor during the relevant times.
According to the Defendants, Mr. Luo indicated that other changes to the
fermentation conditions could result in a change of titres. Mr. Luo spoke about
possible changes to aeration, pressure, the media and agitation, and about
increased familiarity with the process. However, when Mr. Luo’s testimony is
examined, only one explanation – increased agitation – is really offered. And,
I have difficulty with that explanation.
[285]
Exhibit 156 was presented to Mr. Luo during cross-examination. He was
asked what could have explained the increased titres after the plant shutdown.
Mr. Luo testified that a 23% jump could be justified as follows:
A 23 percent jump,
it’s not necessarily like by changing the [strains]. It can also be changed by
optimizing the fermentation conditions of the medium.
. . . There is lots of
factors. It is not only the strain.
As the staff, the
production staff, get more familiarized with the production procedure and
techniques, and then they get improvement on that, it will also change.
So even for the same
strain and the different batches, even for the same strain, the different
batches may also have a different – like, a different result in titre.
[286]
When questioned, Mr. Luo suggested that a change in aeration could improve
production. However, he was unable to point to any change in aeration from the
pre-shutdown runs to the post-shutdown runs.
[287]
Mr. Luo explicitly agreed that there was no change in pressure.
[288]
With respect to a change in medium, Mr. Luo was questioned about the
removal of Amicase. His explanation was bewildering. While he acknowledged that
removing Amicase would probably reduce the titres, he stated that the removal
of Amicase would not necessarily reduce titres. This runs counter to other
testimony on the issue and to common sense.
[289]
At the end of the line of cross-examination on the issue of increased
titres, the only possibly concrete suggestion from Mr. Luo was that increased
agitation during the fermentation could account for the 23% rise in titres.
[290]
Apotex argues that a change in agitation could indeed explain the change
in titres. Apotex first refers to production parameters set out in Exhibit 49 where
a cultivation condition of “agitation between 100 and 130 (depending on DO)” is
set out for AFI-1. In contrast, Apotex notes that the agitation described in the
Batch Records for AFI-4 is closer to 170. Moreover, Apotex points to Exhibit
94, an AFI comparative report where it is reported in section 5.5 that, “for most
of fermentation time, less vigorous agitation is needed for AFI-1 compared to
AFI-4.”
[291]
I give no weight to the information contained in Exhibit 49 referred to by
Apotex. First, the document appears to have been produced no later than spring
of 1995 as part of the transfer of technology related to AFI-1 lovastatin from
AFI to Blue Treasure. I have no idea of what the ultimate instructions followed
were and no explanation of why an agitation of 100 to 130 should be used. I
also observe that the document is made subject to the note that:
This is a preliminary
document and many process parameters including agitation, aeration,
media volumes, cultivation durations, feeding amounts and frequencies etc. are
subject to additional adjustments during the validation runs.
[Emphasis added.]
Beyond the Batch Records, I have no
information on what the actual agitation levels for AFI-1 in production were or
should have been. This document does not assist me.
[292]
I also give no weight to the statement contained in section 5.5 of Exhibit
94. Exhibit 94 is an untitled document that was produced by AFI during the
pre-trial litigation. The document appears to relate to differences between the
AFI-1 and AFI-4 processes. It was referred to by counsel for Merck during
cross-examination of Dr. Sailer. Dr. Sailer described the document as “some
kind of R & D report possibly” and admitted that he had never seen the
document. Although Dr. Sailer was questioned on some of the contents of
this document, I have no idea who is the author of this document or whether
section 5.5 is accurate.
[293]
That leaves Apotex’s analysis of the Batch Records as the only support for
the increased agitation argument. I acknowledge that there appears, in general,
to be an increase in agitation after the plant shutdown. However, I have no
reliable evidence that would support Mr. Luo’s statement that an increase in agitation
explains the dramatic increase in titres.
[294]
This leaves me with no credible explanation for the post-shutdown increase
in titres other than a change in organism. Thus, even if the Batch Records are
accepted as reliable evidence of the production runs, the data with respect to
titres are consistent with the use of the AFI-1 process and not the
non-infringing AFI-4 process. In spite of the removal of Amicase and a decrease
in fermentation duration, Blue Treasure managed to obtain increased titres. The
only explanation consistent with the evidence before me is that Blue Treasure
was using a different micro-organism – Aspergillus terreus.
(5)
Motivation, Means and Opportunity
[295]
Supporting the arguments above, Merck also submits that Blue Treasure had
the motive, means and opportunity to make and ship infringing lovastatin to
AFI.
(a) Motivation
[296]
First, Merck asserts that Blue Treasure had a significant financial
motivation to make infringing lovastatin that it would then sell to AFI. A
number of facts certainly appear to support this argument:
·
the financial difficulties prior to the introduction of AFI-4;
·
the problems with the production of lovastatin using AFI-4; and
·
the inability of Blue Treasure to make a profit at the Blue Treasure Joint
Venture contract price.
[297]
Financial problems at Blue Treasure appear to have existed even before the
production of AFI-4 lovastatin. As we know, Blue Treasure was expected to sell
AFI-1 lovastatin in China or other foreign markets. The Minutes of the 6th
Meeting of the Board of Blue Treasure held August 26-27, 1996 described
the difficult market conditions in China and indicated serious
financial problems at Blue Treasure. At that time, unpaid loans and utilities
bills would leave Blue Treasure with no funds to produce the planned
lovastatin. Dr. Cox, during cross-examination, acknowledged that, at that time,
the outlook was not “very promising”.
[298]
A concern for Blue Treasure from early 1997 was the challenge of selling
into the Chinese market. In a memorandum, dated March 7, 1997, Mr. Zhou
referred to the “less than satisfactory” sales of lovastatin in China and
commented as follows:
In domestic market,
there is a fierce rivalry market of Bulk Lovastatin course price to decline. As
far as we know, ZeJing province Hai Mei Pharm. Offered the price of Bulk
Lovastatin is RMB thirty-two thousand [approximately $3,900 US as confirmed by
Mr. Fowler (T4565)] for one kg. If we do that, we will difficult to subsist.
[Emphasis added.]
[299]
Mr. Fowler, who was responsible for financial matters at AFI during the
relevant times, confirmed that RMB 32,000 would convert to approximately US
$3900. In the absence of any other evidence on the pricing economics, I accept
that Blue Treasure’s “break even” price for the sale of lovastatin was about US
$3900.
[300]
The financial difficulties of Blue Treasure apparently were worsened by
the refusal of AFI to provide higher-yielding strains of Aspergillus terreus.
In the same memorandum referred to above, Mr. Zhou “urgently” requested that
AFI provide it with an improved strain or “super fungus”. Since Mr. Zhou did
not testify at the trial, I can only assume that he believed that a better
strain of Aspergillus terreus would have helped Blue Treasure’s dire
financial situation. AFI did not respond to this request from Blue Treasure.
Dr. Cox acknowledged the refusal:
Q. You had a
contract, after having given them that strain, which obliged you to give them
all of the improvements, correct?
A. Correct.
Q. You were
withholding that, despite your certain knowledge of their financial situation?
A. Yes
Q. And despite
the contractual obligations?
A. If they were
able to sell material with the old strain we might have considered doing that.
There's no point in transferring a high performing strain when they can't sell
the bulk material.
[301]
The decision to send the AFI-4 micro-organism to Blue Treasure was made,
at least in part, to allow Blue Treasure to make non-infringing lovastatin that
it could sell on the Canadian market. In spite of this, the evidence before me
shows that Blue Treasure had some difficulties with producing lovastatin with
the strain of AFI-4 sent by AFI. The result was that the financial picture for
Blue Treasure did not improve.
[302]
Blue Treasure’s experience with AFI‑4 was troubled. Wide
fluctuations in batch-to-batch titres were observed at Blue Treasure prior to March
1998. As acknowledged by Dr. Cox, running a plant economically and efficiently
requires a range of variability within acceptable limits; variability outside
the acceptable range would make it harder to earn a profit. A memorandum, dated
August‑13, 1997, from Mr. Zhou to Dr. Cox highlighted the difficulty:
Because the fermentation
viscosity of new AFI process have a great fluctuation, some equipment in
downstream didn’t suit for the new technology and brought many difficult for
downstream.
[Emphasis added.]
[303]
Blue Treasure also experienced difficulties with some of the AFI-4 strains
provided by AFI. In a memorandum dated October 20, 1997, Mr. Zhou noted that a
replacement strain for L4-42-581 showed only 50% of the productivity and:
Therefore, it can’t be
used for production. Also the strain L4-39-581 was not satisfactory for
production: titers of shake flask test and fermentation showed more than 20%
productivity reduction will occur if Strain L4-39-581 is used compared to that
of Strain L4-42-581. This will certainly affect margin of cost and price.
[304]
Mr. Zhou requested that AFI send “100 frozen vials of L4-42-581 or vials
with similar productivity or higher”. The request for more vials of L4-42-581
was repeated by Dr. Su in an e-mail dated October 19, 1997. As confirmed by Mr.
Primrose, a shipment of seed vials that included 10 vials of L4-42-581 was sent
by AFI to Blue Treasure on August 4, 1998. I have no other evidence of any
response to the request of Blue Treasure for a better AFI-4 strain.
[305]
Apotex asserts that the fact that Blue Treasure was requesting additional
information and a better AFI-4 strain is inconsistent with Merck’s hypothesis
that Blue Treasure was, in fact, using the infringing process after March 1998.
Given that Mr. Zhou, the author of most of the memoranda relied on by Apotex,
did not testify in this trial, I have no way of determining whether Apotex’s
theory is believable.
[306]
As noted above, Blue Treasure was concerned about being able to subsist
with pricing of about US $3900 per kg for Aspergillus terreus
lovastatin. Considerable evidence was produced at trial to demonstrate that the
costs of producing AFI-4 would be higher and that the profit margins even smaller.
[307]
As stated in a lengthy AFI Monthly Scientific Report, dated July 3, 1997,
in which the AFI‑1 and AFI-4 processes were compared, “The calculated
cost of production of AFI-1 was significantly less than the production costs
for AFI-4.” The same document stated that the media ingredients were the single
most significant cost factor.
[308]
The offer to transfer the AFI-4 process to Blue Treasure and to purchase
AFI-4 lovastatin was outlined in a letter dated April 10, 1997 from Dr. Cox to
Mr. Zhou. As confirmed by Dr. Cox, Blue Treasure initially asked for US $6000
per kg. AFI agreed to pay only US $4500 to US $5000, depending on the average
titres. A price of US $4500 would be about $600 to $700 over Blue Treasure’s “break
even” price of US $3900 per kg. However, the margin would have been
significantly eroded by the added costs of the AFI-4 media ingredients and by
having to meet AFI’s 0.2% RC-14 specification. Moreover, over the course of the
batches sold to AFI, the purchase price dropped to US $3300 per kg.
[309]
The result is that, if Blue Treasure had been using the AFI-4 process, it
would have been losing significant amounts of money for each kilogram of
product shipped to AFI. Blue Treasure could not have made a profit by selling
AFI‑4 lovastatin to AFI at the price AFI paid. However, the evidence of
Mr. Fowler was that, once Blue Treasure began to use the AFI-4 process, they
“worked their way out of the financial difficulties and became profitable”.
Quite simply, there is no evidence on the record that explains how Blue
Treasure could have turned a profit using the AFI-4 process.
[310]
I conclude that Blue Treasure had financial motivation to use the AFI-1
process instead of the non-infringing – but more costly – AFI-4 process.
(b) Means
[311]
The ability or means of Blue Treasure to produce the quantities of AFI-1
lovastatin depends, at least to some extent, on the availability of the
appropriate strains of AFI-1.
[312]
Apotex argues that it was not shown that Blue Treasure had such a strain.
Apotex refers to the production master batch records provided to Blue Treasure
as part of the Aspergillus terreus technology transfer. In production,
the strain showed titres generally below the 1.5 g/L level. Not only was the
titre level well below 2.0 g/L reflected in the Batch Records, but the stated
fermentation time was 12 to 15 days – a longer period than reported in any of
the post-March 1998 runs. Merck does not agree.
[313]
Merck submits that Blue Treasure was in possession of a high-producing
strain of AFI‑1 giving it the means to make additional AFI‑1
lovastatin with the very titres reflected in the Batch Records after March 1998
when the average titres were about 2.2 g/L. The reference to this strain is
contained in the Management Report for the 6th Meeting of the Board
of Directors of the Blue Treasure Joint Venture held on August 26-28, 1996. In
that document, at paragraph 1.3, the titres are stated to be 2300 to 2500 mg/L.
Moreover, a document entitled “Historical review of Lovastatin production improvement
Aspergillus sp. and Conioth[y]rium fuckelii strains in R&D
Biology” states that AFI achieved up to 5100 mg/L. When Mrs. Hu was asked whether
Blue Treasure had an AFI-1 strain that produced titres in the range of 2300 to
2500 mg/L, she claimed to be unable to know what the question was about. This
response is surprising given Mrs. Hu’s responsibilities for the security of
various strains of both AFI-1 and AFI-4 and for inoculating the initial
fermentations for each batch.
[314]
With respect to fermentation duration, the evidence of Dr. Cox was that
the AFI-1 strain transferred could be run in nine days.
[315]
Although there appears to be some contradictory evidence, I conclude that
it is more likely than not that Blue Treasure had a strain of AFI-1 that could
have met the titres and fermentation durations reflected in the Batch Records
after March 1998.
[316]
I am persuaded that Blue Treasure had the means to produce lovastatin with
the infringing AFI-1 process.
(c) Opportunity
[317]
The final element examined by Merck is opportunity.
[318]
As described earlier in these reasons, Dr. Jerry Su was sent to Blue
Treasure to investigate certain problems with the initial AFI-4 runs. Dr. Su
arrived in China on August 25, 1997 and returned to Winnipeg on October
31, 1997. Dr. Su expressed his opinion that the runs he observed were carried
out in accordance with the Standard Operating Procedure for AFI-4. However, as
confirmed by Mr. Fowler, after Dr. Su’s departure in October 1997, no one was
ever sent from AFI to assess the production. AFI apparently assumed that Blue
Treasure would follow the instructions that had been provided in spite of
earlier problems.
[319]
AFI also missed another chance to identify the micro-organism used by Blue
Treasure. As confirmed by Ms. Christofalos, with each shipment of product to
AFI, Blue Treasure sent a sample consisting of a mycelial or fungal cake. Dr. Connors
opined that such product could have been subjected to DNA testing for the
presence of the non-infringing Coniothyrium fuckelii. However, the
fungal cake was never tested and was, unfortunately, destroyed during the
course of this litigation.
[320]
Blue Treasure had the opportunity to revert to manufacturing AFI-1 as soon
as Dr. Su left at the end of October 1997.
(6) Blue
Treasure Conduct
[321]
Merck submits that the behaviour of Blue Treasure raises serious doubts
about the Defendants’ allegations that there has been no infringement. One of
the most serious and disturbing stories involves two articles published in the
Chinese Journal of Antibiotics. Merck argues that these two articles are direct
evidence that Blue Treasure used Aspergillus terreus to manufacture
lovastatin, most likely in late 1997. Furthermore, the behaviour of Blue
Treasure’s employee, Mr. Luo, in association with the articles, provides
evidence that Blue Treasure is prepared to go to great lengths to cover up its
infringing activities.
[322]
In early 1999, an article entitled “Optimization of Formula for Lovastatin
Fermentation Medium through Uniform Design” was published in the Chinese
Journal of Antibiotics, February 1999, Vol. 24, No. 1 (referred to as Article
#1). One of the authors of this article was listed as “LUO Dingjun”. Mr. Luo
who appeared as a witness in this trial is the same Mr. Luo who wrote this
article. As set out in the abstract of Article #1, the authors were
presenting a method of optimizing a fermentation medium for producing
lovastatin. In summary terms, the authors conducted experiments to ascertain
the effectiveness of replacing two “extremely expensive” raw materials –
Peptone and Amicase – in the fermentation medium with domestically produced raw
materials. One of the replacement products was a fish powder, giving rise to a
reference, during this trial, to the experiments as the “fish powder
experiments”.
[323]
Article #1, in its abstract, contains the following sentence:
The company [Blue
Treasure] has imported patented lovastatin production technology from Canada, with the
fermentation medium formula being a United States patented (patent
number 5409820) formula.
[324]
The reference to “patent number 5409820” is most likely a reference to
United States Patent No. 5, 409, 820 issued to Apotex Inc. on April 25, 1995.
This patent claims a process for making lovastatin using Coniothyrium
fuckelii. In spite of this reference at the beginning of Article #1, there
is no further reference to Coniothyrium fuckelii. Indeed, the article
explicitly states that the fish powder experiments were conducted on the
“Lovastatin-producing strain Aspergillus terreus BN 270‑001.”
There is no reference in Article #1 to the use of a Coniothyrium fuckelii strain
of lovastatin. The article also describes Guangdong Blue Treasure
Pharmaceutical Co. Ltd. as the company who was producing lovastatin in China and the
company that was interested in the lowering of the costs of the fermentation
medium for lovastatin.
[325]
A second article, entitled “A Study Regarding Lovastatin-Producing Strain
Protoplast Mutation and Fermentation Formula Optimization,” was published in
the Chinese Journal of Antibiotics, October 2000, Vol. 25, No. 5 (referred to
as Article #2). Mr. Luo was also one of the authors of this article. The
purpose of Article #2, as reflected in the abstract, was to produce a
high-yield strain of lovastatin “by mutating lovastatin-producing strain
protoplasts” and optimizing the fermentation culture medium formula. Once
again, the authors listed, as the strain used for their experiments,
“lovastatin-producing strain Aspergillus terreus (BN 270-053)”. In
Article #2, there is no mention of the US Patent.
[326]
Dr. Luo testified that the shake flask experiments (referred to in Article
#1 and #2) were performed at Blue Treasure’s microbiology lab.
[327]
These articles raise a strong inference that, at the time of the
experiments, Blue Treasure was producing lovastatin from Aspergillus terreus
at its facility.
[328]
Mr. Luo was asked about the two journal articles during his
examination-in-chief. In response to a question on the purpose of publication,
Mr. Luo responded that:
Getting this paper
published is mainly for the technicians being able to get certified. Basically
it's a kind of when you come to a profession new to get certified,
certification.
[329]
Mr. Luo was unable to remember clearly when the experiments were run, but
he thought that they were carried out in 1997 “between summer and fall of that
year”. Mr Luo further testified as follows:
Q. Where was
these experiments conducted?
A. These
experiments are done at Blue Treasure's micro‑biology laboratory.
A. Specifically
what strain was used I can't remember so clearly, but looking through this
article probably I used number 4 strain.
Q. Why does it
say Aspergillus terreus was used?
A. Because at
that time due to the confidentiality reasons I don't want to leak the secrets
about AFI‑4 strain.
[330]
I find Mr. Luo’s testimony related to these two articles to be astounding.
First, it is simply incomprehensible that a group of scientists would fabricate
a journal article to hide the identity of the Coniothyrium fuckelii
fungus. Beyond a bald assertion, Mr. Luo presented no cogent reasons as to why
the use of Coniothyrium fuckelii would have been confidential.
[331]
I also question Mr. Luo’s testimony with respect to when the experiments
were conducted. His lack of memory is surprising. Mr. Luo has known about the
significance of these articles for some time and yet was unable to recall the
dates of the experiments. He could have refreshed his memory in a number of
ways before testifying. Instead, he expressed the view that the experiments
were done in the period of time before the use of the AFI-4 process. His faulty
memory, in my view, is consistent with fabrication, by Mr. Luo, of what
organism was being used at Blue Treasure in the time leading up to the
publication of the articles.
[332]
The two articles contain other references that discredit Mr. Luo’s claims.
First, the strain used for the articles was described as BN 270-053. In the
opinion of Dr. Lasure, this was a reference to the 53rd isolate of Aspergillus
terreus BN-2-70. This is the same strain obtained from AFI as AFI‑1
(Lasure Expert Report, Exhibit 48, para. 242).
[333]
The second area of disconnection relates to the ability of the strains of Coniothyrium
fuckelii used in AFI-4 to sporulate. The evidence of Dr. Cox was to the
effect that, in his experience working at AFI, the strain of Coniothyrium
fuckelii transferred to Blue Treasure had lost the ability to produce
spores during fermentation:
Q. All right.
You know that at some point Coniothyrium fuckelii ceased sporulating at AFI?
A. Yes, I
believe so.
Q. That makes
it harder to prepare and maintain consistent seed banks from generation to
generation?
A. A little
bit, yeah.
Q. That
cessation of sporulation is a property of the organism?
A. Not of the
organism, no, it was an effect of the mutagenic strain improvement program that
AFI performed.
Q. The non
sporulating character became intrinsic to Coniothyrium fuckelii as a result of
that program, is that fair?
A. That
strain of Coniothyrium fuckelii, yes.
Q. Right.
A. It's not an
inherent characteristic of the species.
Q. Got it. Never
a problem that was encountered with Aspergillus terreus?
A. No.
[Emphasis added.]
[334]
In spite of the information that the AFI-4 strain sent to Blue Treasure
could not sporulate, in Article #2, at section 1.2.3, the authors explicitly
report as follows:
Investigation of
strain-producing capacity: The spores on the mature PDA slants are washed, inoculated
onto a rice culture medium, and cultured for 7 days, yielding mature rice
spores.
[335]
In sum, Mr. Luo’s explanation about why he lied in the two articles is not
believable. Mr. Luo fabricated his testimony before this Court to cover up
the use of Aspergillus terreus at a time when Blue Treasure was supposed
to have been exclusively using the AFI-4 process. The better explanation for
these two articles and Mr. Luo’s testimony is that the fish powder experiments
were carried out on lovastatin using AFI-1. I find that, on a balance of
probabilities, the experiments were carried out with the AFI-1 process (that
is, with Aspergillus terreus). It follows that the two articles support
Merck’s claims that Blue Treasure was using Aspergillus terreus to make
lovastatin and thus infringing the '380 Patent.
C. Conclusion on Blue Treasure
Circumstantial Evidence
[336]
Most of the foregoing evidence could be characterized as circumstantial
evidence and requires that I draw inferences to reach the conclusion that, on a
balance of probabilities, there was infringement.
[337]
With respect to the drawing of inferences, Apotex argues that I must avoid
relying on alleged inferences that are no more than speculation. I agree.
[338]
Obviously, there is a difference between reasonable inference and pure
conjecture. The Federal Court of Appeal provided the following comments in an
appeal of an immigration judicial review, in Minister of Employment and
Immigration v. Satiacum (1989), 99 N.R. 171 at p. 179, [1989] F.C.J. No.
505 (QL)(F.C.A.):
The common law has
long recognized the difference between reasonable inference and pure
conjecture. Lord Macmillan put the distinction this way in Jones v. Great
Western Railway Co. (1930), 47 T.L.R. 39 at 45, 144 L.T. 194 at 202 (H.L.):
The dividing line
between conjecture and inference is often a very difficult one to draw. A
conjecture may be plausible but it is of no legal value, for its essence is
that it is a mere guess. An inference in the legal sense, on the other hand, is
a deduction from the evidence, and if it is a reasonable deduction it may have
the validity of legal proof. The attribution of an occurrence to a cause is, I
take it, always a matter of inference.
In R. v. Fuller
(1971), 1 N.R. 112 at 114, Hall J.A. held for the Manitoba Court of
Appeal that "[t]he tribunal of fact cannot resort to speculative and
conjectural conclusions." Subsequently a unanimous Supreme Court of Canada
expressed itself as in complete agreement with his reasons: [1975] 2 S.C.R. 121
at 123, 1 N.R. 110 at 112.
[339]
In the case before me, I am satisfied that I am not relying on conjecture
or mere guess; rather, the deduction from the totality of the evidence that
Blue Treasure was manufacturing lovastatin using the infringing AFI-1 process
is reasonable and logical.
[340]
Given the extent of the evidence before me and the lack of alternative,
reasonable explanations for much of that evidence, I find that Merck has
satisfied its burden to show that it is more likely than not that, from March 1998,
Blue Treasure was manufacturing lovastatin using the infringing AFI-1 process.
The lovastatin was then delivered to AFI for ultimate sale into the Canadian
market. This constituted infringement of the '380 Patent.
[341]
This conclusion stands without any reference to the DNA evidence. That is,
regardless of whether the DNA evidence presented by Merck constitutes reliable
evidence of direct infringement, I find that any amount of lovastatin
manufactured and sold in Canada that used lovastatin produced by Blue Treasure
after March 1998 infringed the '380 Patent.
[342]
The DNA evidence is dealt with in section VIII of the Reasons.
D. AFI Batch CR0157
[343]
Merck no longer claims that all of the lovastatin produced in Winnipeg by AFI
infringed the '380 Patent. However, Merck continues to assert that one batch of
lovastatin – referred to as CR0157 – contained lovastatin that was made using
the process protected by the '380 Patent.
[344]
Batch CR0157 consisted of 12.57 kg of USP-grade lovastatin. It was
manufactured at AFI in Winnipeg and sent directly to the attention of Dr. Barry Sherman at
Apotex Inc. on December 2, 1996. It was the first batch of AFI-4 lovastatin
shipped to Apotex Inc.
[345]
The batch history reflects that CR0157 was the final crystallization of CR0151
which, in turn, was a recrystallization of CR0149. CR0149 was the
crystallization of three extraction batches: CR0117, CR0145 and CR0144. The
“genealogy” of batch CR0157 was confirmed by Dr. Sailer, during his examination-in-chief,
and is depicted in Figure 1 below.
Figure 1
[346]
The essence of Merck’s argument is that CR0157 was a combined batch of
non-infringing and infringing lovastatin. The key evidence relied on by Merck
is:
(1)
AQA 94 Batch Records (referred to as AQA 94) show that CR0157 was “AFI#1,4
FLAGGED contains top secret material” which would indicate it was a mixture of Aspergillus
terreus and Coniothyrium fuckelii lovastatin;
(2)
Some of the batches that ultimately resulted in batch CR0157 (specifically
CR0151, CR0117, CR0144 and CR0149) show significant alterations that increased
the weights of lovastatin from those initially recorded.
(3)
The DNA testing carried out by Dr. Davies demonstrated that tablets of Apo‑lovastatin
tested positive for Aspergillus terreus and negative for Coniothyrium
fuckelii.
[347]
In my view, for the reasons that follow, neither of the first two
arguments is persuasive on its own. However, the direct evidence of Aspergillus
terreus in tablets made from the CR0157 batch was uncontradicted and
provides direct evidence of infringement.
[348]
As part of the trial record, the Defendants produced AFI’s batch records.
These records were obviously produced contemporaneously with the fermentations
to which they refer and were kept in AFI’s normal course of business. In
contrast to the Batch Records for Blue Treasure, the AFI batch records are
acceptable as business records. In other words, the AFI batch records are prima
facie proof of the facts stated therein.
[349]
However, it does not follow that every document, produced by AFI, that
looks like a business record meets the strict criteria to be admitted into
evidence as a business records. This is particularly true with the document
referred to as AQA 94 and relied on by Merck.
[350]
AQA 94 is a one-page document consisting of a list of 56 batch numbers
with brief notations. The entry of interest is at line 11:
CR0157 AFI#1,4 FLAGGED
contains top secret material.
[351]
The document was introduced to Dr. Cox who was unable to identify it or to
provide any testimony that could have assisted the Court. Ms. Christofalos was
asked about the document during her examination-in-chief.
Q. Are you able
to identify this document?
A. This looks
like one of our internal, well, it is one of our internal record‑keeping
documents where we just have made an index of a bunch of documents that is
going into a box.
Q. Do you know
what purpose was made for it?
A. Just for
filing purposes.
Q. Do you know
when it was made?
A. We started
this exercise late 99. I would, that's my recollection, late 99.
Q. Can you tell
the Court who made it?
A. Could have
been any number of people, most likely our records clerks.
[352]
Although Ms. Christofalos admitted that she did not prepare the document,
she expressed her view that the list was merely an identification of boxes of
documents. Further, Ms Christofalos provided what, in my view, was a reasonable
explanation of why there was the reference to “AFI#1,4” in the document:
·
The person making this list would have looked at the first page of the
actual Interim Production Batch Record and would only have seen “Product Name:
lovastatin USP” and “Product Part Number: 6000”.
·
Part number 6000 applied to lovastatin made with either AFI-1 (At)
or AFI-4 (Cf), thus leading to the entry listing both.
·
The reason why the error was never corrected was because these lists were
not control documents.
[353]
With all respect to Merck, I fail to see how this document of dubious
origin can be strong circumstantial evidence of an infringing product in a
batch of lovastatin. Dr. Cox was clearly unaware of AQA 94. Ms. Christofalos’s
explanation was simply that this was an index of documents. She did not prepare
the list and was unsure of who could have prepared it. The list of documents
was only made in 1999 – more than two years after batch CR0157 was
manufactured. Unlike the AFI batch records, AQA 94 does not qualify as a
business record. I do not accept it for the truth of its contents – in
particular, that CR0157 was a mixture of infringing and non-infringing
lovastatin. In light of the lack of connection between AQA 94 and the
“genealogy” of CR0157, I give no weight to this document.
[354]
The only remaining argument of Merck (other than the DNA evidence) relates
to some apparent anomalies surrounding the weight of the three preliminary
batches that ultimately became part of CR0157. In making this argument, Merck
refers to the batch records for three antecedent downstream batches: CR0151, CR0117
and CR0144. In each case, a document that forms part of the batch records shows
a manual change to the originally-recorded weight of the material obtained. In
each case the weight was increased. In Merck’s submission, the augmentation is
consistent with the addition of infringing lovastatin to the batch.
[355]
In response, the Defendants attempted to provide an explanation for these
discrepancies through the evidence of Dr. Sailer and Ms. Christofalos.
[356]
Dr. Sailer clarified that the antecedent weights going into CR0157 cannot
simply be added up, as it is a weight loss process.
Q. Right. Well
again, I don't have any errors to complain about there either. The only point
I guess I'm conceding, and I want to make sure I understand that I have it
right, is that we can't simply add together these three numbers?
A. We can
because the batch 151 was generated from batch CR0149 and those two errors
which we see in the material coming to batch CR0149, which you said it's plus
11.2 plus 1.2 kilograms, generated certain quantity.
Q. Right.
A. As a
starting material for the next batch 151. During the next batch when they
tried to discharge it they have to correct it because they discharge extra 3.8
kilograms, which actually doesn't matter at all because if they didn't
discharge it, we will dissolve it anyway and mix it with the 11.11 kilograms to
generate 14.8 kilograms of product, which was starting material for the batch
157 of the final product.
Q. All right.
[357]
Further, at a later date, extra lovastatin was found that had been
captured in the filter, which was not initially observed by the operators. It
was subsequently recorded. Dr. Sailer points out that the crossed out weights
were the result of mistakes by inexperienced technicians.
Q. All right.
So let's go to tab three, page 82. Have you got that page?
A. Yes.
Q. And I'm
seeing a table one crude Lovastatin obtained, do you see that?
A. Yes, I do.
Q. And I see
the first listing is CR0117?
A. I can see
it.
Q. But I see
the net weight is 3.8 kilograms.
A. Yes.
Q. What
explanation, if any, do you have for the Court for that?
A. I have
similar explanation. This is first batch which was filtered on this equipment,
NF 1, and this is what they found when actually they dismantled the screen,
they dropped the bottom and they found there was still product there. I know
it should be recorded here but as I said it was first batch. And yeah, we
didn't have really experience. We were quite new crew there, so yeah, it's
missing, missing note that additional 1.2 kilogram was found on the screen.
[358]
Finally, I observe that the corrections to the records were made some
eight days after the original entry. When asked about the delay, Dr. Sailer was
unable to provide a response.
[359]
While the explanations given by Dr. Sailer and Ms. Christofalos could
provide a sufficient answer for the discrepancy, they are only that - an
explanation or hypothesis. Their answers do not explain why there was an
eight-day delay in changing the document or provide a fulsome explanation of
how the process of making lovastatin would “lose weight”.
[360]
In sum, I am left with unanswered questions on the make-up of batch CR0157.
Since the batch was never tested by Apotex, I have no direct evidence that the
batch was not Aspergillus terreus lovastatin. However, in the submission
of Merck, I do have direct evidence that tablets from the CR0157 batch were
made from infringing lovastatin. Merck makes this argument based on the DNA
testing results of Dr. Davies. That evidence is considered in section VIII of
these Reasons.
VIII. Infringement
– the DNA Evidence
A. Introduction
[361]
Merck’s case on the direct evidence of infringement rests
on the DNA testing carried out on behalf of Dr. Julian Davies in his laboratory
at the University of British Columbia (referred
to as the Davies Lab). Merck relies on the Expert Report and testimony of Dr.
Davies to assert that there is direct evidence that lovastatin manufactured by
Blue Treasure and lovastatin contained in AFI batch CR0157 contained Aspergillus
terreus, thereby infringing the '380 Patent.
[362] Dr. Davies received and tested three samples of lovastatin, consisting of:
(a) three vials of white powder and one vial of brown powder, purporting to be
samples from Blue Treasure provided through the services of Mr. Ted Kavowras;
and (b) Apo-lovastatin tablets made from AFI batch CR0157.
[363] Two
sets of experiments were carried out at the Davies Lab on the same samples –
the first in 2003 and another in 2007. In 2003, the experiments were performed
by Karen Lu, a senior technician for the Davies Lab. In 2007, Grace Yim, a Ph.D
candidate in the Davies Lab, assisted Karen Lu in performing the experiments.
[364] In
2003, the experimental procedure involved a number of steps: DNA was extracted
from the samples; specific primers diagnostic of the fungi region in DNA were
chosen from the published literature; the primers were used to amplify the DNA
in a nested polymerase chain reaction (nested PCR); the amplified DNA was
subjected to gel electrophoresis; and then the gel was stained and analyzed
under ultra-violet (UV) light for the presence of DNA. The UV light analysis
highlighted the presence of bands which correlated to 130 base pairs (bp).
These bands were determined to be a positive “hit” for Aspergillus terreus
DNA. The bands were extracted from the gel and sent to a lab which specializes
in DNA sequencing. The Davies Lab compared the results of the DNA sequencing to
the sequence for Aspergillus terreus DNA that is located in the database
at the National Institute for Biotechnology Information. This comparison
confirmed that the DNA “hit” in the lovastatin samples was, in fact, Aspergillus
terreus DNA.
[365] In
2007, the Davies Lab repeated the experiments with a modified procedure and
tested for the presence of both Aspergillus terreus and Coniothyrium
fuckelii DNA. Again, a positive result was only found for Aspergillus
terreus.
[366]
The combination of the experiments performed in 2003 and
2007 resulted in 13 positive findings of Aspergillus terreus DNA, and no
positive findings of Coniothyrium fuckelii DNA. Based on these results,
Dr. Davies opined that the fungus responsible for producing the lovastatin in
the samples tested by the Davies Lab was Aspergillus terreus.
[367]
The first general and critical consideration is that the
burden that falls on Merck is one of a balance of probabilities. That is, Merck
succeeds if I am satisfied that it is more likely than not that the samples tested
contained Aspergillus terreus DNA.
[368]
A number of issues arise with respect to the DNA evidence
presented through Dr. Davies:
1.
Is there a nexus between the Blue Treasure samples tested
and the Blue Treasure lovastatin that is alleged to infringe the '380 Patent?
2.
Should the results of the testing in the Davies Lab be
rejected because they were not reproducible by Dr. Poinar?
3.
Does the failure of Dr. Davies to find C. fuckelii
DNA in the tablets from batch CR0157 support Apotex’s position that his DNA
evidence is unreliable?
4.
Is “ancient DNA” the same as, or analogous to, the DNA that
would be found in the samples tested by Dr. Davies?
5.
Should the results of the testing in the Davies Lab be
rejected because the Aspergillus terreus DNA found in the Davies Lab was
the result of contamination by exogenous DNA?
6.
Should the weight to be given to Dr. Davies’s Expert Report
and his opinions be reduced due to:
a)
the inability to describe elements of the experiments
reported in his Expert Report;
b)
the incomplete reporting of the experiments conducted in
the Davies Lab;
c)
the failure to disclose the majority of tests relied on; or
d)
the failure to use negative extraction controls?
[369]
To this list, I would add the issue of the weight to be
given to the opinions of the experts put forward by Apotex – Drs. Gilbert,
Poinar and Taylor. Should the narrow expertise of these witnesses reflect on
the weight to be given to, or the relevance of, their opinions?
B. Nexus
between the samples tested and the allegedly infringing lovastatin
[370] An assessment of the DNA evidence in this case necessarily begins with the
source of the samples that were tested in the Davies Lab. The presence of Aspergillus
terreus in the samples only establishes infringement if the samples were
obtained: (a) from lovastatin that was manufactured by Blue Treasure during the
relevant time period; or, (b) from Apo-lovastatin tablets whose source was AFI batch
CR0157.
[371] I begin with the three vials of white powder and one vial of brown powder
allegedly obtained from Blue Treasure. In 2000, a law firm representing Merck
hired Mr. Ted Kavowras to obtain samples of Blue Treasure lovastatin.
Mr. Kavowras has an investigative consulting firm based in China
called Panoramic Consulting. Most of his investigations are done undercover.
Mr. Kavowras has considerable experience obtaining evidence used in civil litigation.
[372] Mr.
Kavowras testified that, using the alias of “Mr. Garcia”, he approached Blue
Treasure to obtain lovastatin samples. In October 2000 and January 2001, he
obtained samples from an employee of Blue Treasure. Samples, from two batches,
were delivered in a sealed package with a corresponding certificate of analysis
for each of the two samples - recorded as batch numbers 200012016 and
200012015. During his testimony at trial, Mr. Kavowras confidently identified
the exhibits produced as the samples that he obtained from Blue Treasure. The samples were transported in carry-on luggage that accompanied Mr.
Kavowras and were stored in his office in a secure location.
[373]
Merck presented Ms. Giuliani to speak to the delivery chain
from the hands of Mr. Kavowras to Dr. Davies. The samples were delivered to Ms. Giuliani who stored them in the company vault, and arranged for their delivery
to Dr. Davies. Dr. Davies testified that the samples he received were in
“perfect” condition.
[374]
The evidence presented by Merck shows that the chain of
evidence was intact and that the samples delivered to Dr. Davies were in an
unaltered form.
[375]
Apotex does not take issue with the transmittal of the
samples to Dr. Davies. However, the problem is that these samples were taken
from Blue Treasure between 2000 and 2001. There is no clear evidence that these
samples were samples of the Blue Treasure lovastatin manufactured during the
relevant time period from 1997 to 1999. Merck has failed to persuade me that
the samples are representative of the lovastatin made, sold and distributed by
Apotex Inc. Without the link of the samples taken by Mr. Kavowras to the lovastatin
produced by Blue Treasure during the 1997 to 1999 period, I am unable to
conclude that any DNA evidence from Dr. Davies can establish infringement.
[376]
The situation with respect to the Apo-lovastatin tablets is
different. There is no dispute that the source of the tablets tested was AFI batch
CR0157. Thus, if I conclude, on a balance of probabilities, that the CR0157
tablets tested in the Davies Lab contained the DNA of Aspergillus terreus
that was not the result of contamination of the samples, infringement has been
demonstrated.
[377]
Moreover, if I am wrong in my conclusion with respect to
the lack of nexus in the Blue Treasure samples, the question before me would be
whether the evidence establishes that those Blue Treasure samples contained the
DNA of Aspergillus terreus that was not, on a balance of probabilities,
the result of contamination. If it does, infringement has been demonstrated.
C. Reproducibility
of the testing in the Davies Lab
[378] Apotex argues that I should reject the testing done in the Davies Lab because
Dr. Poinar was unable to reproduce his results.
[379] In
addition to being asked to provide his opinion on Dr. Davies’s work, Dr. Poinar
was retained by counsel for Apotex to design and carry out experiments with the
same materials that were tested in the Davies Lab. Dr. Poinar tested 12 lovastatin
samples made by AFI or Blue Treasure. For control purposes, Dr. Poinar’s
experiments included lovastatin samples that were known to have been produced
with Aspergillus terreus. According to Apotex, Dr. Poinar used an
extraction method that was essentially the same as Dr. Davies’s protocol. Dr.
Poinar concluded that: “both A. terreus and C. fuckelii DNA were
absent in all samples tested”. By way of apparent explanation, Dr. Poinar
observed that (Poinar Expert Report, Exhibit 135, p.15 “conclusion”):
The sample processing and/or extraction steps are too
lossful to permit detection of trace amounts of DNA from the source fungus
(neither A. terreus or C.fuckelii was detected in samples of
known origin).
[380] In
other words, Dr. Poinar concluded that it is impossible to obtain detectable
amounts of DNA from pharmaceutical fungal products using the experimental
methods employed by Dr. Davies.
[381] Apotex
relies on the evidence of Dr. Poinar to draw the conclusion that any positive
results observed by Dr. Davies were as a result of contamination, not
endogenous DNA.
[382] It
appears that Dr. Poinar approached his experiments with the opinion that it is
unlikely that DNA would survive (Poinar Expert Report, Exhibit 135, para. 62):
… a careful examination of the entire procedure for the
fermentation, extraction, recrystalization and purification of lovastatin makes
it unlikely that DNA would survive this procedure.
During his oral
testimony, he restated this view:
So it is surprising,
although certainly not impossible, that DNA would actually then survive that.
[383]
Given Dr. Poinar’s initial starting bias, is there any
doubt that his experiments would fail to find DNA? It appears to me that there
is a real risk that Dr. Poinar brought a confirmational bias to his laboratory
work. In other words, I question whether Dr. Poinar carried out the experiments
to confirm his thesis that DNA would not be present, rather than with an open,
independent mind.
[384]
Even putting aside my concern about Dr. Poinar’s approach
to the experiments, I have serious problems with his experimental methods and
results.
[385]
The first concern relates to Dr. Poinar’s experiences with
the Aspergillus terreus assay. On this issue, Dr. Gilbert, accepted as
an expert in the area of microbial genetics and microbiology, was of great assistance to the Court. During cross-examination, Dr. Gilbert
was shown certain of the data from Dr. Poinar’s experiments. In particular, he
was shown two sets of qPCR data. Without knowing (as we do now) whether the
information came from the Coniothyrium fuckelii assay or the Aspergillus
terreus assay, he was asked to comment on various aspects of those data.
The first series of slides consisted of data from the Coniothyrium fuckelii
assay. Dr. Gilbert described the various melting curves as “very
good-looking” and “pretty nice”. With respect to the Coniothyrium fuckelii
series of slides, he agreed that the data reflected an “experiment in which the
PCR reactions [are] well behaved and as expected” and that the reaction was
“running well”. Dr. Gilbert acknowledged that the standard curve for the Coniothyrium
fuckelii assay was typical of “a PCR reaction that is running well and
predicatively as expected”. The gel samples shown to Dr. Gilbert were
described as “pretty much a textbook example”.
[386]
The situation changed dramatically when Dr. Gilbert was
shown the Aspergillus terreus data. For example, whereas the efficiency
for one of the Coniothyrium fuckelii slides was described as 93.5%, the
efficiency of a standard curve for an Aspergillus terreus assay was only
70%. Dr. Gilbert opined that the experimenter was “having problems with his
quantification” and that “possibly”, he was having problems with the primers,
reactants or inhibition, problems that were not apparent in the Coniothyrium
fuckelii assay.
[387]
Overall, Dr. Gilbert agreed that, even apart from
quantification, the stochasticity, randomness and unexpected results were an
indication that the A. terreus experiment did not run as well or as
smoothly as the Coniothyrium fuckelii assay. He also acknowledged that
the data supported this conclusion.
[388]
The necessary corollary of Dr. Gilbert’s opinion is that
Dr. Poinar’s results are reliable regarding the absence of C. fuckelii but
that they cannot be used as a basis to rule out the presence of A. terreus.
[389]
The next problem with Dr. Poinar’s work is that he did not
employ the same methods that were carried out in the Davies Lab. Using a single
round of PCR, Dr. Poinar was unable to get any hits with the A. terreus
primers at the 0.1 copy level. However, he was able to get a hit using Dr. Davies’s
nested PCR reaction. Despite the possibility that it would confirm Dr.
Davies’s results and despite success at the 0.1 copy level with Dr. Davies’s
protocol, Dr. Poinar did not attempt to reproduce or run a single sample
through Dr. Davies’s entire nested PCR protocol. Finally, Dr. Poinar
agreed with the following premise put to him in cross-examination:
Q. You can't,
based on your experiments, rule out the possibility that if you had done his
experiment, reproduced his experiment, you might have found terreus?
A. No.
[390]
Based on the evidence before me, I cannot conclude that Dr.
Davies’s work is not reproducible.
D. Failure of
Dr. Davies to find C. fuckelii DNA in the tablets from batch CR0157
[391]
As noted above, Dr. Davies found A. terreus DNA in
the tablets from AFI batch CR0157 but did not find any C. fuckelii DNA
in the tablets. On its face, this appears to be inconsistent with Merck’s
theory that batch CR0157 was likely a mix of Coniothyrium fuckelii
lovastatin and Aspergillus terreus lovastatin. In final argument, Apotex
stated that this alleged inconsistency “is
fundamentally fatal to any reliance upon the Davies [Lab] testing”, and
described its position as follows:
[Dr. Davies] did not
find CF when he should have, but he found AT when he should not have. The AT
could only, in that scenario, have been exogenous AT.
....
The other alternative
on CR0157 is that it is a mixture. Some AT lovastatin was thrown into the mix,
but much of it, most of it, I believe, even on that scenario, was CF
lovastatin.
So on that scenario,
it contains both CF and AT lovastatin. Dr. Davies should have detected AT and
he should have detected CF. He still didn't detect CF under that scenario,
which calls into question his testing and calls into question whether the DNA
survives, and indicates that the AT that he did find, again, must be exogenous
DNA, because he didn't find the CF.
[392]
The position of Apotex is based on what I believe is an incorrect
assumption. A failure to find a particular micro-organism through DNA testing
is not determinative that the micro‑organism is not present. On the other
hand, assuming appropriate laboratory procedures and no contamination, a
finding of a specific micro-organism is strong evidence that the micro‑organism
is present in the sample. This was acknowledged by Dr. Poinar during his cross‑examination:
Q. . . . If you want
to say to someone, Look, in my opinion, DNA is absent from this sample, the
actual -- the scientifically correct proposition is, always, DNA is absent from
this sample to a state of detection limit; correct?
A. That's
correct.
Q. But the flip
side of the coin for the positive finding, it is not the same in science. In
science you can say, I found terreus in this sample. I don't know how much was
there, but I amplified it?
A. Hmm hmm.
Q. So to that
extent, the positive finding is different from what you need to prove the
negative?
A. As long as
all the proper precautions are put into place. So had all of the extraction
controls and PCR controls been there, then the possibility would exist; that's
correct.
[393]
For example, we know that Dr. Poinar did not find any A.
terreus DNA in any of the samples that he tested. Yet, we also know that there
were lovastatin samples provided to Dr. Poinar – in addition to the samples in
issue – that clearly contained Aspergillus terreus DNA. Nevertheless,
Dr. Poinar was unable to confirm the presence of this micro-organism. As
discussed above, I have rejected Dr. Poinar’s presumption that DNA cannot be
found in such samples. It follows that his conclusion that Aspergillus
terreus DNA was not present in the samples is incorrect; Aspergillus
terreus DNA was present and Dr. Poinar simply did not use adequate experimental
methods to find the DNA.
[394]
Applying this analysis to Dr. Davies, I conclude that the
failure of Dr. Davies to find Coniothyrium fuckelii, even if it existed
in batch CR0157, does not mean his finding with respect to Aspergillus
terreus cannot be relied on. It certainly does not follow, as
asserted by Apotex, that the DNA found by Dr. Davies in AFI batch CR0157 must
be exogenous.
[395]
A large caveat must be introduced at this point. I have
rejected Apotex’s reliance on either Dr. Poinar’s inability to find Coniothyrium
fuckelii or Aspergillus terreus or Dr. Davies’s failure to find Coniothyrium
fuckelii DNA in the CR0157 tablets. However, that does not necessarily mean
that I will accept Dr. Davies’s conclusions. As stated by Dr. Poinar, “all the
proper precautions” must be in place before one can rely on the results of the DNA
testing. That brings me to the key issue with respect to the DNA evidence
relied on by Merck: Is the DNA evidence reliable? Stated differently, were
“proper precautions” used in the Davies Lab? Any discussion of this question
must begin with the opinions of the experts put forward by the Defendants in
response to Dr. Davies’s DNA evidence.
E. DNA
evidence and the Apotex Experts
[396]
Apotex’s response to the DNA evidence rests on the
criticisms of Dr. Davies’s work provided by three experts – Drs. Gilbert,
Poinar and Taylor.
[397]
Dr. Taylor was qualified to provide opinions to the Court
as an expert mycologist, and microbiologist with particular expertise in fungal
DNA evolution and fungal DNA PCR amplifications. Dr. Taylor opined on Dr.
Davies’s Expert Report and DNA testing results, particularly with respect to
the possibility of contamination in the Davies Lab.
[398]
Dr. Gilbert was qualified as an expert in the analysis of
low copy and degraded DNA. His expert testimony and report deal with the issue
of infringement. He explains the problems with the use of ancient or degraded
DNA (referred to as ancient DNA). Dr. Gilbert was asked to examine Dr. Davies’s
laboratory notebooks, and comment on whether the conclusions generated were
justified by the methods used.
[399]
Dr. Poinar was qualified as an expert in the extraction and
characterization of low copy and degraded DNA. Dr. Poinar reviewed the
experiments performed by the Davies Lab, and (discussed above) attempted to
replicate the results through testing of his own. Dr. Poinar was asked to
answer two general questions: first, whether or not the experimental design of
Dr. Davies matched the rigour necessary for an ancient DNA or low template
sample project; second, whether Dr. Davies’s data support his hypothesis that Aspergillus
terreus DNA was found in samples of lovastatin from the Defendants.
[400]
As a general rule of evidence, witnesses may not give
opinion evidence, but may only testify as to the matters within their
knowledge, observation and experience. An exception to this rule applies to
expert witnesses. Experts are necessary to assist the Court in scientific
matters that would not normally be within the knowledge of the judge. However,
before accepting the opinion evidence, the trier of fact must determine the
admissibility of the evidence in accordance with four criteria: relevance,
necessity in assisting the trier of fact, the absence of an exclusionary rule
and a properly qualified expert (R v. Mohan, [1994]
2 S.C.R. 9, 114 D.L.R. (4th) 419).
[401]
In this case, Drs. Taylor, Gilbert and Poinar appear to
satisfy three of the four criteria. However, given the narrowness of their experience
and the focus of their opinions on ancient DNA, I seriously question the
relevance of their opinions. This is true for all three experts, although more
so with respect to Dr. Poinar and Dr. Gilbert. While the level of my concern
does not lead me to conclude that all of their opinion evidence is
inadmissible, it does reflect on the weight that should be given to their
opinions. The key problem exists with the characterization of the DNA tested by
Dr. Davies as ancient DNA.
(1) What is Ancient DNA?
[402] The first question is: what is ancient DNA? On the record before me, there
is no clear definition of this term. However, each of the Defendants’ experts
provided insights on how to understand what ancient DNA is:
·
Dr.
Poinar stated that (Poinar Expert Report, Exhibit 135, para. 2):
The study of ancient DNA is the retrieval and meticulous
characterization of DNA sequences from samples which are assumed to be heavily
degraded, in low copy numbers and typically stemming from forensic, fossil,
sub-fossil, archeological, and palenontological remains.
·
Dr.
Gilbert stated that (Gilbert Expert Report, Exhibit 130, para. 8):
I have been involved in the analysis of a range of
materials that contain degraded DNA, and/or low copy number modern DNA. These
include ancient bone and tooth material, often dating back tens of thousands of
years in age, hair shaft and root, nail, horn, skin (both tanned into leather
and dried), mummified soft tissues, feather, eggshell, ancient plant seeds and
leaves, formalin and Bouin’s solution fixed soft tissues, historic blood
samples, feces, urine, soil, ice and honey.
·
Dr.
Taylor stated that (Taylor Expert Report, Exhibit 124, para. 25) :
This scarcity of DNA in the lovastatin samples makes
analysis of DNA from these samples akin to analysis of DNA from archeological
samples, the field of research referred to above as “ancient DNA”. A key
complication of ancient DNA research is the high likelihood of contamination of
the sample with DNA from modern sources of by DNA that has been amplified by
PCR from modern DNA, unless the necessary precautions are taken.
[403]
There is a clear gap in these opinions – they fail to
provide a comprehensive analysis of how DNA was degraded in the process of making
a pharmaceutical product, or even how to compare the DNA in ancient DNA to DNA
derived from a pharmaceutical product. Without this evidence, I am unable to
compare the evidence presented on ancient DNA to evidence presented by Dr. Davies.
[404]
Dr. Taylor made a sweeping statement in his report that
“the scarcity of DNA … is akin to archaeological samples”. In the absence of a
comprehensive analysis, this is neither a convincing statement nor a scientific
one. What is the basis for his knowledge that the DNA in lovastatin is scarce?
Where is the evidence that shows us that these two types of DNA can be
considered analogous? Without something further, I do not see how it can be
logical to compare the level of DNA available from ancient
bone and tooth material to DNA derived from a
pharmaceutical product.
(2) Is DNA derived from a pharmaceutical product degraded or
fragmented?
[405]
The next step is to consider the condition or state of DNA
derived from a pharmaceutical product. During the trial, there was some
controversy over the semantics of whether the DNA derived from the lovastatin
samples, and tested by Dr. Davies, was “fragmented” or “degraded”. Apotex
argues that Dr. Davies specifically described the DNA that was the subject of
his experiments as “degraded DNA”. Although this is correct, Dr. Davies
provided a sufficient explanation of what was meant by the use of this term in
his Expert Report.
[406]
During his oral testimony, Dr. Davies described the process
whereby the fungal DNA in lovastatin is fragmented during industrial
processing. He explained that DNA is released by the cells which expose the DNA
to shearing forces in the fermentation machine. This was not contested. The
issue was whether these shearing forces would cause “fragmentation” or
“degradation”. Dr. Davies was asked about this during cross-examination:
Q. You just
told me a few moments ago that there would be fragments or, the word that you
didn't like to use, degradation; would I be right that you would agree with
this, that although some of the DNA survives in the pharmaceutical process
involved here, there, unquestionably, would be some DNA that would be degraded
during the process?
A. In the case of
a fermentation process, I think I mentioned this before, during fermentation
and particularly at the end, cells begin to LYSE, they break and they release
DNA. That DNA is going to be sheared; it's going to be exposed to shearing
forces of the big stirrers that are used in the fermentation. And shearing will
break DNA.
For one
thing, Aspergillus DNA often has a lot of protein bound to it which might
stabilize it, but we planned our experiments with the expectation that the DNA
was going to be sheared. It would never shear down to almost nothing.
[407]
In his report, Dr. Davies uses the word “degraded” when
referring to the DNA which was the subject of his experiments. However, during
oral testimony, Dr. Davies clarified that he equates the word “degraded” to
“fragmented” – which is the word he should have used in his Expert Report. Dr. Davies
explained that fragmentation is a normal process that occurs to all DNA which
is outside a cell. He did not intentionally refer to “degradation” as
synonymous with the process that results in ancient DNA.
Q. You talk
about the difficulties your lab experienced, you say:"I underscored the
difficulty in obtaining DNA from pharmaceutical samples."And then you
should read it, but go to the sentence that begins, "although some DNA
survives the process."
A. Yes.
Q. It says:
"There is
unquestionably some DNA that is degraded during the process."Do you accept
that as correct, that's your view?
A. I do.
THE COURT: Is that
because you equate the word degraded with fragmented?
THE WITNESS: Yes.
This is different.
THE COURT: Let me get
here to paragraph 63. When you use the word "degraded" in paragraph
63, do you mean fragmented?
THE WITNESS: I mean a
very general process.
THE COURT: Is there
something beside fragmentation that you mean by the word degraded in that
paragraph?
THE WITNESS: Yes,
Your Honour. There are enzymes, chemicals, in the reaction in the fermenter
that could chemically modify the DNA.
[408]
Dr. Davies goes on to opine that there is no clear evidence
that the A. terreus DNA is “degraded” (in terms of the ancient DNA
reference) during the fermentation process.
Q. Going back to my question, do you not agree with
this, that a though some DNA survives the fermentation, purification process,
that there, unquestionably, would be some DNA that is degraded during the
process?
A. If we buy degradation, we're talking about
chemical reactions, and I would say yes, it's possible.
Q. Not unquestionable?
A. I don't know, you see. Nobody's ever looked.
You've got a fermenter with things DNA has never seen and you're stirring it
up.I don't know anybody is going to look to see what the DNA would be like and
how it was broken. We didn't look at the fermenter. We only looked at the
powder.
[409]
In my view, Dr. Davies’s explanation is clear as to why DNA
derived from a pharmaceutical process could be considered to be “fragmented”
but not “degraded”.
[410]
So, what does this mean to me? Apotex’s experts assume that
the DNA tested by the Davies Lab was “degraded” in the same manner that ancient
DNA is degraded. I do not agree. The only relevant evidence presented before
me on this issue was that of Dr. Davies. I conclude that the DNA tested by Dr.
Daves is considered to be “fragmented” but not “degraded”.
(3) Can one
compare how DNA derived from a pharmaceutical product is
fragmented
to how DNA from “ancient DNA” is
degraded?
[411]
After considering the evidence on what is ancient DNA, I will
now consider whether DNA derived from a pharmaceutical product that is
“fragmented” can accurately be referred to as ancient DNA. I conclude that it
cannot.
[412]
In my view, it does not logically flow to compare DNA from
“ancient bone and tooth material, often dating back tens of thousands of years”
to DNA from Aspergillus terreus which has been processed into a
pharmaceutical product. This is true especially in light of understanding that
the subject DNA is not “degraded” in the same manner.
[413]
Without specific expert evidence on this, it is not logical
to compare the two. The Defendants’ experts provide no relevant evidence on
this point. However, Dr. Davies explicitly states that one cannot compare
ancient DNA to DNA derived by a pharmaceutical process.
Q. Tell me what
you understand the words "ancient DNA" to mean to at least those that
use the expression?”
A. As I
understand it, it is DNA being isolated from remnants of earlier civilization,
animals, from earlier stages, things of that type, and looking for DNA from
those samples and trying to identify them. So the sequencing of some old human
genomes recently. I don't know if that's called ancient DNA or not.
Q. In your
description of it, would I be correct that the DNA one has in the context that
you've given me for in ancient DNA, is DNA that is either low copy number DNA
or degraded DNA?
A. Depends how
you define degraded.
Q. How would
you define degraded?
A. I think
there are two kinds of degradation.
Q. What are
they?
A. One is that
the DNA can be broken to give very low molecular weight fragments which makes
it more difficult to do a PCR.
. . .
Q. On the last sentence, the process of cleaving
that you talked about in that sentence, that causes, according to your
sentence, further degradation, not a new but further suggestion that the
earlier process that you described, the purification steps earlier, causes
degradation, and that leads to further degradation through the cleavage from
the enzymes, correct?
A. That is correct. What I should have said is that
the DNA is really modified and broken up in a horrible way, and there would
be very little intact chromosomal DNA but there would be fragments, and some of
those would not work in PCR because they'd be modified further. That's the difference
between this and ancient DNA, and things like that.
Q. I don't know what you mean by that. What do you
mean by ancient DNA?
A. It's DNA from million year old remains and things
of this type.
Q. You're suggesting that that DNA is not fragmented
and not degraded and not broken up?
A. It's everything. It's even worse. So the
techniques used for ancient DNA are really very specific, and it's not the same
as this.
Q. Because it's even more extreme and ancient DNA
versus what you encountered here?
A.Yes, cosmic rays have been acting on it for a long time.
But as I think I mentioned right at the beginning, one of the biggest
problems with what they call ancient DNA is the question of desiccation, and
desiccation causes quite severe damage to DNA.
[Emphasis added.]
[414]
Dr. Davies provides his expert opinion that DNA that has
been hit with cosmic rays for thousands of years – DNA that scientists did not
think even existed until a few short years ago – cannot be compared with DNA
that has been processed into a pharmaceutical product. I agree with him.
(4) Are the
opinions of the Defendants’ experts relevant to fragmented DNA from
pharmaceutical
products?
[415]
Having concluded that ancient DNA is not comparable to DNA
derived from a pharmaceutical product, I turn to a consideration of whether the
opinions presented by the Defendants’ experts extend beyond the field of
ancient DNA. In other words, can the opinions of any of the experts assist the
Court in gaining a different understanding of the DNA testing from fungal
pharmaceutical compounds than what was provided by Dr. Davies?
[416]
In my view, none of the experts has provided me with the
necessary link between ancient DNA and the DNA in issue. Apotex has failed to
satisfy me that the evidence of these experts is relevant to the issues before
me.
[417]
Dr. Waldo Taylor was the only expert (other than Dr.
Davies) whose expertise extended beyond the field of ancient DNA. During his
examination-in-chief, he stated the following:
Q. I just want
to ask you about the expression "ancient DNA". Can you describe what
you understand that to mean?
A. Sure. As I
mentioned, when the PCR reaction became available it was possible to examine
DNA and biological specimens, or really any specimen, where DNA was wanting,
where it was scarce. So that field has grown and it encompasses people who
study ancient DNA, who study DNA that survives in the fossil specimens, things
like Neanderthal bones or organisms in amber. It includes forensic studies
where there's a little amount of DNA associated with some social interaction
that you'd like to study. It includes the study that we're concerned with
today, trying to identify the organism that produced a pharmaceutical compound.
[Emphasis added.]
[418]
However, during his cross-examination Dr. Taylor acknowledged
that he was not fully aware of “the whole story” of how the lovastatin samples
tested in the Davies Lab were treated or prepared.
[419]
Dr. Taylor’s opinions are seriously undermined by his
admission that he lacks understanding of how the relevant pharmaceutical
compounds have been prepared or treated. It is inadequate for an expert to
explicitly state that the subject DNA is comparable to ancient DNA when he does
not know the scientific basis for this comparison.
[420]
Neither Dr. Gilbert nor Dr. Poinar offered their opinions
on how, if at all, fragmented DNA from a pharmaceutical compound could be
treated as ancient DNA.
[421]
According to Dr. Gilbert, there are millions of scientists
in the world who have access to, use or rely upon PCR data; yet, there were
only a “handful of people on the earth” who are qualified to undertake the kind
of ancient DNA analysis Apotex promotes. It would have been more helpful to the
Court (and relevant to the issues of this case) if Apotex had presented DNA
experts who operated outside this “handful of people on the earth”.
[422]
In sum, I have difficulty with the opinions of the Apotex
experts. Their opinions are based on an assumption that Dr. Davies’s work and
opinion could be evaluated against the same standards, protocols and norms as
would be used by the “handful of people on the earth” with an expertise in
ancient DNA. I am not satisfied that this assumption is correct. This problem
informs my evaluation of the criticisms levelled at Dr. Davies’s work. The
Apotex experts are providing their opinions through a very narrow prism that is
only marginally relevant – at best – to the issues before me.
F. Contamination
[423]
As we know, Dr. Davies had 13 hits for Aspergillus terreus DNA from
three different samples. In their arguments, the Defendants raise many, many
problems with Dr. Davies’s opinion, his testimony at the trial and the
procedures and results from the Davies Lab. I have considered all of their
concerns. The most serious allegation is the risk that the Aspergillus
terreus DNA found in the samples tested was the result of contamination by
exogenous Aspergillus terreus. In other words, the Defendants submit
that the risk of contamination in the Davies Lab was so high that I cannot
accept the DNA testing evidence of Dr. Davies as reliable.
[424]
As all of the experts would agree, contamination is always a serious risk
in DNA testing. I acknowledge the risk. The question for this Court is whether
the risk of contamination makes it more likely than not that Dr. Davies’s Aspergillus
terreus DNA findings are false.
(1) Dr.
Davies
[425]
Dr. Davies was clearly aware of the risk of contamination. Dr. Davies
addressed the issue of laboratory contamination on several occasions during his
testimony and in his Expert Report. Dr. Davies stated (Davies Expert
Report, Exhibit 55, para. 77):
I do not consider contamination
to be a likely explanation to account for the presence of Aspergillus
terreus. For example, if our laboratory was contaminated, I would have
expected to see Aspergillus terreus amplicons in the negative controls.
However, this was not the case.
[426]
Dr. Davies cautioned that there is always a possibility of a contamination
while carrying out a PCR reaction with DNA. However, during his oral testimony,
he insisted that, “[the Davies Lab] took every possibility to avoid
contamination. We’re not forensic scientists, but we can work clean.”
[427]
Dr. Davies likened the scenario of contamination in the subject
experiments to “the chance that lightening would hit you.” He explained that
his two investigators, Grace Yim and Karen Lu, were extremely careful workers
and precautions were taken to prevent contamination, including:
·
working in a sterilized area to kill bacteria and fungi that could be a
source of contamination;
·
autoclaving the micro-pipets and tips to sterilize and prevent
cross-experiment contamination;
·
operating
in a UV-radiated, stainless steel biosafety hood to purify the incoming
air and allow for easy cleaning of the hood; and
·
utilizing experiments with zero DNA controls, and following the protocol
that the experiment would be repeated if contamination was found in the zero
DNA control.
[428]
Dr. Davies was convinced that contamination was not responsible for the
results of these experiments. He opined that the fact that the Davies Lab never
found Coniothyrium fuckelii or any other fugus of the genus Aspergillus
(other than terreus) as a contaminant, was evidence to support his
assertion that contamination was not responsible. Dr. Davies observed that it
was hard to believe that, if contamination had been present in the experiments,
it would only be found in the form of Aspergillus terreus.
[429]
Lastly, as Dr. Davies most poignantly stated, “DNA doesn’t fly” or “float
around in the air”.
[430]
Against the backdrop of Dr. Davies’s expert testimony, I turn to the
evidence of the experts produced by Apotex. In general, as discussed above, the
Apotex experts brought a very narrow perspective to the question of
contamination. All three imposed incredibly high standards on the work of DNA
laboratories and began with the assumption that Dr. Davies was dealing with DNA
that was akin to ancient DNA that might be found in the 1000-year old remains
of extinct animals. Quite simply, these experts imposed an overall standard on
the DNA testing performed by the Davies Lab that is not reasonable in the
circumstances. Two experts – Dr. Taylor and Dr. Gilbert – were particularly stubborn
in their opinions on contamination in the Davies Lab.
(2) Dr.
Taylor
[431]
Dr. Taylor equated the analysis of the tested lovastatin to an “analysis
of DNA from archaeological samples” (Taylor Expert Report, Exhibit 124, para.
25). For the reasons expressed earlier, I am of the view that this assumption
is flawed, or unsupported by the record.
[432]
Directly following from this assumption, Dr. Taylor expressed the view that
a “substantial risk of contamination” could come from two sources: (a) cultures
of Aspergillus terreus which were used by Merck to produce lovastatin;
or, (b) Aspergillus terreus DNA that was PCR amplified in the Davies Lab.
[433]
Dr. Taylor concluded that (Taylor Expert Report, Exhibit 124, para. 28):
If Dr. Davies did in
fact detect Aspergillus terreus DNA, then his own laboratory practices
very likely caused his samples to become contaminated with such DNA from Aspergillus
terreus cultures or from DNA that been PCR amplified from Aspergillus
terreus DNA.
[434]
During cross-examination, Dr. Taylor stated that the most likely source of
contamination was Aspergillus terreus DNA from PCR amplifications in the
Davies Lab. This theory was discussed with Dr. Gilbert during cross-examination.
Dr. Gilbert explained why the amplified material posed more of a risk; it is
because the “PCR part” would “be more concentrated” with DNA. However, as the
evidence shows, the strains of both Aspergillus terreus and Coniothyrium
fuckelii were present in the Davies Lab at the relevant time. Dr. Gilbert
also confirmed that the risk of contamination from the control DNA from Aspergillus
terreus or Coniothyrium fuckelii was equivalent, even though the DNA
concentration would differ. Yet, Dr. Davies found Aspergillus terreus DNA
in 13 experiments but never documented a single instance of Coniothyrium
fuckelii. How could it be that amplified Aspergillus terreus DNA
would cause contamination but not the Coniothyrium fuckelii DNA? Dr.
Gilbert repeatedly refused to accept the possibility that zero findings of Coniothyrium
fuckelii DNA should reassure the experimenter that contamination by Aspergillus
terreus was not probable. I find his denials to be unhelpful and, frankly,
illogical. Thus, Merck submits, Dr. Taylor’s theory does not withstand
examination. I agree. Zero findings of Coniothyrium fuckelii
demonstrates the weakness in the theory that contamination would, more likely
than not, occur from Aspergillus terreus DNA that was amplified by PCR
in the Davies Lab.
[435]
Another theory of contamination presented by Dr. Taylor was the
possibility of airborne spores. I do not accept this as likely in the Davies Lab.
The use of a biosafety hood with purified air would dramatically reduce the
possibility of such contamination. Further, Dr. Taylor acknowledged that he had
never studied the ability to amplify from a single spore of Aspergillus
terreus and admitted that it was speculative to conclude that a single spore
would amplify by PCR.
[436]
A final concern of Dr. Taylor relates to his reading of the various “gels”
that recorded the experimental results. Dr. Taylor noted a “milestone” on the
gels. Dr. Taylor observed the following (Taylor Expert Report, Exhibit 124,
para. 33):
. . . Dr. Davies’
laboratory records from 2007 confirm that DNA was amplified from the lovastatin
samples using coniothyrium fuckelii specific primers, thus indicating the
presence of coniothyrium fuckelii DNA in the sample. However, Dr. Davies did
not mention this in his affidavit. Accordingly, Dr. Davies’ failure to
acknowledge coniothyrium fuckelii as the potential source organism, based on
his erroneous assertion that no coniothyrium fuckelii DNA was found, was not
warranted.
[437]
In other words, Dr. Taylor was of the opinion that Dr. Davies had positive
findings of Coniothyrium fuckelii that were overlooked or never
reported. I agree that an unreported instance of Coniothyrium fuckelii
would be a critical contention with respect to the possibility of contamination
and also with respect to Dr. Davies’s conclusion that there was no Coniothyrium
fuckelii DNA found in any of the samples tested. However, examination of
Dr. Taylor’s statement and observations by Dr. Gilbert and Dr. Poinar
diminish the significance of this observation.
[438]
During his cross-examination, Dr. Taylor acknowledged that his conclusion
that this was Coniothyrium fuckelii was speculative; he agreed that the
first amplification of DNA using a C. fuckelii primer may have been
Coniothyrium fuckelii DNA or “may be OJ Simpson’s DNA”. Dr. Poinar
agreed that, based on the data, one could not conclude that the band observed
by Dr. Taylor was a band of Coniothyrium fuckelii DNA. When Dr.
Gilbert was asked about the existence of any bands that were the correct size
for Coniothyrium fuckelii using the Coniothyrium fuckelii
specific primers, Dr. Gilbert responded that there were “zero” such bands.
[439]
Viewed as a whole, I do not find that Dr. Taylor’s opinions regarding contamination
are supported by the record.
(3) Dr.
Gilbert
[440]
Dr. Gilbert provided his expert opinion that “contamination presents a
serious challenge to the data generated by Dr. Davies’s team”. I do not believe
that anyone would disagree with this statement. However, Dr. Gilbert continues
to express his view that, “as a result, his results cannot be viewed as
reliable.”
[441]
The first problem that I have with Dr. Gilbert’s opinion is that it is
directed to his field of specialization – the fields of ancient DNA and
forensic genetics. In his Expert Report, Dr. Gilbert refers to samples that
“were millions of years old”. His entire opinion is premised on the assumption
– unexplained, in my view – that Dr. Davies was working with degraded DNA that
was comparable to ancient DNA. As discussed above, I am not persuaded that this
is a meaningful comparison.
[442]
Moreover, the contamination theory posited by Dr. Gilbert (and Drs. Taylor
and Poinar) is that exogenous Aspergillus terreus resulted in the
positive findings by the Davies Lab. Thus, unless the experts can identify a
credible source of the exogenous Aspergillus terreus DNA, the theory
does not stand up to scrutiny. Upon cross-examination, Dr. Gilbert acknowledged
that there was no evidence that contaminants, if they existed in the Davies Lab,
were Aspergillus terreus.
[443]
Further, how has Dr. Gilbert measured or defined the term “reliable”? Can
the possibility of contamination explain away every one of the 13 “hits”? Does
the risk of contamination make it more likely than not that Aspergillus terreus
does not exist in the samples? Or, is contamination merely a possibly?
[444]
In sum, I am not persuaded that the risk of contamination in the Davies Lab
rises to the level where, on a balance of probabilities, I cannot rely on the
results obtained and opined on by Dr. Davies.
G. Other criticisms of Dr.
Davies’s opinion
[445]
The Defendants levelled a number of other criticisms at the work of Dr.
Davies. Specifically, the Defendants submit that the following principles must
be followed to determine whether the Court should ignore the expert evidence of
Dr. Davies:
·
Where an expert’s conclusion is not appropriately explained and supported,
it may properly be given no weight by the trier of fact (Backman v. Canada (1999),
178 D.L.R. (4th) 126 at para. 34, 246 N.R. 309 (F.C.A.)).
·
Likewise, for a Court to accept an expert’s opinion, the trier of fact
must know the facts and/or assumptions upon which the expert has based his or
her opinion (Johnson & Johnson v. William H. Rorer, (Canada) Ltd.
(1980), 48 C.P.R. (2d) 58 at para. 7, [1980] F.C.J. No. 200 (QL)
(F.C.T.D.)).
·
Where those facts turn out to be inaccurate or incomplete, the weight
given to an expert’s opinion may be significantly reduced (Misik (E.) v. M.N.R.,
[1993] 1 C.T.C. 2360 at p. 2373, [1993] T.C.J. No. 13 (QL) (T.C.C.)).
[446]
In support of their position that the opinion of Dr. Davies should be
rejected, the Defendants point to “spurious bands” seen on some of the “gels”.
These bands, in the view of the Defendants, demonstrate that Dr. Davies’s
opinion was inaccurate.
[447]
In principle, I agree with the Defendant’s argument that an expert should:
(1) appropriately explain their conclusions; (2) inform the Court of the
underlying facts and/or assumptions; and (3) provide accurate and complete
information. If unable to do so, it should be expected that the weight of their
opinion will be significantly reduced.
[448]
However, I believe that the Defendants are misapplying these principles to
the opinions given by Dr. Davies.
(1) Lack
of knowledge
[449]
I agree with the Defendants that there were aspects of the experiments
performed in the Davies Lab that Dr. Davies was unable to describe to the
Court. However, the critical question is whether his knowledge of the technical
aspects of the experiments was fundamental to his expert opinion. I do not
think it was.
[450]
If one accepts that someone is an expert in their field (as the Defendants
did in the case of Dr. Davies), the parties must provide the Court with a clear
understanding of what their expertise is. A problem arose in this case because
of a misunderstanding between what constitutes being an expert in the science
of microbiology and an expert in performing the technical experiments of
microbiology. There obviously was confusion of what was expected of Dr. Davies.
Dr. Davies obtained his Ph.D in 1956, and reached professor emeritus status in
1997. He is unquestionably an expert in his field. One does not rise to that
level at a Canadian university without having outstanding accomplishments in an
area of expertise. However, it is not surprising that someone like Dr. Davies,
who is in charge of a laboratory with many students and technicians, is not
personally performing the experiments and may not be knowledgeable on all of
the technical aspects of the experimental procedure. Obviously (and somewhat
critically of Dr. Davies), I am of the view that Dr. Davies should have
prepared better for his testimony. However, unless the lack of knowledge goes
the heart and substance of Dr. Davies’s opinion, I would not be persuaded to
reject that opinion.
[451]
The Defendants agreed, after reviewing his Expert Report, and before his
oral testimony, that Dr. Davies was an expert in microbiology and microbial
genetics. That expertise includes the use of DNA techniques for studying
microbes. That expertise – and not the minutia of laboratory procedures – was
the focus of his opinion. I do not agree with the Defendants that Dr. Davies’s
lack of knowledge of some the technical aspects of the experiments should
necessarily lead this Court to the conclusion that the evidence is unreliable.
(2) Incomplete
Report & Lack of Disclosure
[452]
The Defendants argue that Dr. Davies was incorrect and incomplete in his
report of, and conclusions as to, the majority of the experiments disclosed. I
agree that Dr. Davies’s notebooks and conclusions were incomplete. However, the
determinative question is whether the missing information is fundamental to his
expert opinion. I do not believe it was.
[453]
The Defendants refer to the following incidents of incompleteness:
·
In 2003, experiments 13, 19 and 20 are disclosed. The inference is that at
least 17 other experiments were not disclosed. Thus, 17 of 20 or 85% of all
experiments were not disclosed.
·
In Karen Lu’s 2007 work, experiments 7, 8, 18, 20, 22, 23 and 24 are
disclosed. The inference is that at least 17 other experiments were not
disclosed. Thus, 17 of 24 or approximately 71% of all experiments were not
disclosed.
·
In Grace Yim’s 2007 work, experiments 32, 33, 35 and 36 are disclosed. The
inference is that at least 8 other experiments are not disclosed. Thus, 8 of 12
or approximately 67% of all experiments were not disclosed.
·
In total, 42 or 54 experiments, or 78% of all experiments were not
disclosed.
[454]
The Defendants submit that Dr. Davies provided an incomplete picture of
the experiments performed by only providing selected pages for litigation. The
Defendants allege that Dr. Davies erred by providing only the pages that “would
bear on the conclusion”, and not all of the pages that were relevant. The
Defendants further argue that Dr. Davies provided an erroneous explanation for
withholding the notebook pages because “the non-disclosed experiments did not
bear on the conclusion ‘in any significant way’”. I do not agree with the
Defendants that this was an erroneous explanation.
[455] During
cross-examination, Dr. Davies testified that the reason he did not disclose all
of the pages relating to all of the experiments performed in his lab was that
this was not “routine” procedure for a scientist. When preparing evidence to
support a publication, it is “routine” for a scientist to provide only the
pages that support the conclusion that the scientist has opined. There was no
malice or bad faith in Dr. Davies’s lack of disclosure.
Q. You deprived a reader of coming to a different
conclusion about the sufficiency or the relevance of the experiments included and
not included, correct?
A. No, I don't think so. May I make the point that
when one publishes a scientific paper, you have to publish a logical sequence
of events, and you publish the results, and you then interpret the results
based on the experiments? This was not a scientific paper. Things were much
simpler here than that. I believe we gave you a continuum. We selected some
experiments. Some of them were experiments that didn't work that well, but you
got an idea as to what was going on. I don't see anything wrong with that.
[456]
During cross-examination, Dr. Davies was presented with a chart, prepared
by counsel, representing the missing information. Throughout the
cross-examination, I was unable to observe any missing notes or details that
would have had a material impact on the overall opinion of Dr. Davies. When
asked about the chart during cross-examination, Dr. Poinar stated that all
documents need to be put forward. However, he also acknowledged that, in this
case, the data depicted in the chart meant little, if anything, to the
conclusions being drawn by Dr. Davies.
[457]
I do not agree with the Defendants that the decision of Dr. Davies to
exclude certain pages must result in a conclusion that his evidence is
unreliable. What is important is that the opinions of the expert not be
contradicted or otherwise weakened by any missing information. By not putting
forward all of the notebook pages, Dr. Davies ran the risk of not being able to
substantiate his opinions. However, in this case, as admitted by Dr. Poinar,
the missing information is not relevant to the overall conclusion. Accordingly,
while I would have preferred that Dr. Davies had provided a more complete
picture of the experiments performed, I do not consider that his opinion to be
fatally flawed by his failure to do so.
(3) Unexpected
results
[458]
The Defendant’s argue that Dr. Davies obtained unexpected results and
therefore his work cannot be relied upon. At the heart of their argument is the
presence of “spurious bands” that only became visible upon dramatic enlargement
of a photograph of the gel that was stained and analyzed under UV light for the
presence of the PCR amplified DNA
[459]
The Defendants argue that the experiments of Dr. Davies were flawed because,
upon investigation of the laboratory notebooks, extra bands could be seen on
the gel pictures. The Plaintiffs, on the other hand, argue that any evidence
involving “photoshopped” gels should not be considered or given any weight by
this Court. I agree with the Plaintiffs on this point.
[460]
When presented with the greatly enlarged photographs, Dr. Davies commented
that the gel photographs were manipulated by the Defendants. Dr. Davies stated:
Your photo shopping is
up to some amazing magnification … people do this to public papers sometimes
and I think it’s unrealistic. It’s not what you see on the actual gel. This is
not the actual gel. This is a doctored gel in my opinion … You can find many
bands on PCR gels that you would not see by eye. You would not see by other
detections if you photo shopped them up in some way. I am just very concerned
about this, and I find it difficult to draw what I would consider to be sound
conclusions
[461]
I agree with Dr. Davies. The original photographs of the gels are highly
technical and – frankly – confusing to this judge. How can I be certain that
taking those already confusing depictions and enlarging them to this
extraordinary scale does not bear a risk of introducing photographic ghosts,
shadows or other images that did not originally exist? I have no evidence that
such distortion of the original evidence would not introduce errors or
unreliable results.
[462]
Even if I were to accept that there are faint bands that were not seen in
the original gels or referred to by Dr. Davies, none of this evidence accounts
for Dr. Davies’s findings of Aspergillus terreus in the lovastatin
samples.
[463]
Overall, the presence of the spurious bands, or the impact of the
photoshopped pictures, does not lead me to the conclusion that the results of
the experiment were flawed.
IX. Infringement
– Conclusion
[464]
My overall conclusion is that Dr. Davies’s testing results are reliable
and credible evidence that the lovastatin samples tested in his laboratory
contained Aspergillus terreus DNA. I have rejected the contamination
theory of Apotex.
[465]
I am satisfied that the lovastatin tablets tested
originated from AFI batch CR0157. The DNA evidence satisfies me that, on a
balance of probabilities, the Defendants infringed the '380 Patent with
the manufacture and sale of any product from AFI batch CR0157.
[466]
The situation with respect to the Blue Treasure samples is different. As
noted earlier in these reasons, I am not persuaded that Merck has demonstrated
a nexus between the samples tested in the Davies Lab and the allegedly
infringing product. Thus, I do not accept the DNA evidence as direct evidence
of infringement by the lovastatin produced by Blue Treasure. Nevertheless, the
remaining evidence presented by Merck with respect to the Blue Treasure
lovastatin strongly supports the conclusion that there was infringement. This
is discussed earlier in these reasons. If I am wrong with respect to the lack
of nexus, then I would conclude that the DNA evidence is evidence of direct
infringement, a conclusion that would strengthen – but not change – my earlier
finding of infringement.
X. Validity
A. Introduction
[467]
Having determined the issues of claims construction and infringement, I now
turn to the issue of validity. In their counterclaims, each of the Defendants
assert that the '380 Patent is invalid. If they are correct, there will be no
question of infringement; the Defendants cannot infringe an invalid patent.
[468]
In an infringement action, the patentee benefits from the presumption of
validity (s. 45 of the Patent Act and s. 43(2) of the Patent Act
currently in force). Thus, Apotex bears the burden of proving, on a balance of
probabilities, that the '380 Patent is invalid. As of the close of argument in
this trial, the following remained as Apotex’s grounds of invalidity:
·
All of the disputed claims, except for claims 3 and 6, are invalid because
they are overly broad.
·
Certain of the claims are invalid because they could not demonstrate
utility. In particular:
○
claims 1, 2, 5 and 13 to 15 demonstrate a lack of utility, in that not all
of the compounds included in the claims are useful in fulfilling the promise of
the '380 Patent; and
○
the utility of the claimed subject matter, including pharmaceutical
utility, demonstrates a lack of sound prediction, in that the inventor could
not predict, as of the Canadian filing date, that the compounds claimed would
fulfil the utility promised by the '380Patent.
·
Claims 13 to 15 are invalid because of prior use or anticipation by the
existence of lovastatin in a traditional Asian product known as “Red Yeast Rice”.
·
The claims are invalid because Merck was not the first to invent
lovastatin.
B. Overbreadth
[469]
Apotex argues that the claims of the '380 Patent, other than claims 3 and
6, are “overbroad” because they claim processes that were not invented by the
named inventors. In their submission, the invention, as claimed, comprises numerous
strains or species which were not tested or evaluated by the inventors to
determine, as of the priority date, whether they were capable of producing the
compounds.
[470]
A patent which claims more than what has been invented can be found to be
invalid as being overly broad. The concept of “overbreadth” has been referred
to in a number of cases before our Court. In support of its position, Apotex
relies on the case of Biovail Pharmaceuticals Inc. v. Canada (Minister
of National Health and Welfare), 2005 FC 9, [2005] F.C.J. No. 7 (QL) [Biovail].
In Bioval, above, at paragraph 61, Justice Harrington described the
notion of “covetous claiming” as follows:
If the inventor claims
more than he should, he loses everything.
His fences must be
clearly placed in order to give the necessary warning and he must not fence in
any property that is not his own. [Thorsen P. in Minerals Separation North
American Corp. v. Noranda Mines Ltd., [1947] Ex.
C.R. 306 at page 52, as quoted in Free World Trust, supra, at
para. 14]
[471]
In Biovail, above, the patent in issue (a patent for a controlled-release
pharmaceutical tablet) disclosed only one sustained-release mechanism, referred
to as an osmotic process, using a compound known as HPMC as the carrier. The
generic company was using a different mechanism and compound – a hydrogel
process using a compound known as HPC as the carrier. Justice Harrington’s key
finding was that the patent was not infringed because the inventors only
contemplated the osmotic process, and not the hydrogel process, which was a substantially
different process. However, in the alternative, Justice Harrington considered
that, if the claims were to be construed to include the hydrogel process, “the
patent was invalid for covetous claiming” (Biovail, above, at para. 60).
[472]
Biovail, in my view, does not support Apotex’s submission on
overbreadth. Unlike the patent in Biovail, the '380 Patent’s claims and
disclosure make reference to all producing species of Aspergillus terreus.
The question of whether the inventors could extrapolate their laboratory
results from the specific samples tested is a question of sound prediction and
not of overbreadth.
[473]
Apotex also relies on Eli Lilly Canada Inc. v. Apotex Inc., 2008 FC
142, 63 C.P.R. (4th) 406 [Eli Lilly Raloxifene (FC)], aff’d 2009 FCA 97,
78 C.P.R. (4th) 388, a decision in respect of a patent that
claimed the use of raloxifene in the treatment of osteoporosis and bone loss.
At paragraphs 179-182, Justice Hughes discussed the assertion that the claims
were overly broad. Key to his conclusion that the claims were overly broad was
the “disconnect” between the disclosure and the claims in the patent. The
disclosure limited the osteoporosis and bone loss to that without the adverse
effects of estrogen therapy. In all but one of the claims in issue, there was
no limitation to the use of the drug for bone loss due to estrogen-related
causes. Thus, Justice Hughes concluded that all but one of the claims in issue
were overly broad. The issue of overly broad claims was not overturned by the
Court of Appeal.
[474]
In my view, Eli Lilly Raloxifene (FC), above, does not support an
argument that the claims of the '380 Patent are overly broad. In the patent
before me, the disclosure is consistent with the claims in issue. Contrary to
the submissions of Apotex, the disclosure of the '380 Patent is not limited to
the two strains of Aspergillus terreus that were used in the experimentation
by Merck that lead to the invention.
[475]
In brief, on Apotex’s claim of overbreadth, I conclude that this argument,
on the facts of this case, is more properly a question of sound prediction - in
particular, both the factual basis for the prediction and the sufficiency of
the disclosure.
C. Utility
(1) General Principles
[476]
Apotex submits that the '380 Patent is invalid on the grounds that:
1.
the impugned claims lack actual utility, in that certain strains of Aspergillus
terreus have been shown to be unable to produce any of the claimed
compounds; and
2.
as of the relevant date, the inventors could not soundly predict that the
claimed compounds would have the utility promised by the '380 Patent.
[477]
Section 2 of the Patent Act defines an invention as something that
is "new and useful". From this comes the concept of
"utility".
[478]
A number of principles associated with the law of utility are well
established in the jurisprudence. To begin, the overarching concept is that, as
of the relevant date, there must have been a demonstration of utility of the
invention or, lacking that, a sound prediction of utility based on the
information and science available at the time of the prediction (see, for
example, Merck & Co. v. Apotex Inc., 2005 FC 755, 41 C.P.R. (4th) 35
at para. 121; Pfizer Canada Inc. v. Apotex Inc., 2007 FC 26,
306 F.T.R. 254 at paras. 36-40, aff'd 2007 FCA 195, 60 C.P.R. (4th) 177,
leave to appeal to SCC refused, [2007] S.C.C.A. No. 371 (QL), 381 N.R. 399
(note)).
[479]
Apotex bears the burden on this issue of validity. To demonstrate lack of
utility, Apotex must show "that the invention will not work, either in the
sense that it will not operate at all or, more broadly, that it will not do
what the specification promises that it will do" (Consolboard,
above, at p. 525).
[480]
For many patents in the pharmaceutical field, the inventors will not yet
have demonstrated that the invention “works”, as of the relevant date. In such
cases, the inventors rely on the concept of “sound prediction”. The doctrine of
sound prediction can be relied upon by an inventor to justify patent claims
whose utility has not actually been demonstrated, but can be soundly predicted
based upon the information and expertise available (Apotex Inc. v. Wellcome
Foundation Ltd., 2002 SCC 77, [2002] 4 S.C.R. 153 at para. 56 [Wellcome
AZT]). A party challenging the utility of a patent based on sound
prediction must demonstrate that the prediction was not sound or that there is
evidence of a lack of utility. As stated in Wellcome AZT, above, at
paragraph 56:
If a patent sought to
be supported on the basis of sound prediction is subsequently challenged, the
challenge will succeed if, per Pigeon J. in Monsanto Co. v.
Commissioner of Patents, [1979] 2 S.C.R. 1108, at p. 1117, the prediction
at the date of application was not sound, or, irrespective of the soundness of
the prediction, "[t]here is evidence of lack of utility in respect of some
of the area covered".
[481]
The relevant date is the date of the filing of the Canadian patent
application (Ramipril I (FC), above, at paras. 88-96). For the '380
Patent, that date is June 11, 1980.
[482]
Where the specification does not promise a specific result, no particular
level of utility is required - a "mere scintilla" of utility will
suffice (H.G. Fox, The Canadian Law and Practice Relating to Letters Patent
for Inventions (4th ed., 1969), at p.153). However, as stated in Consolboard,
above, where the specification sets out an explicit "promise",
utility must be measured against that promise (see, for example, Pfizer
Canada Inc. v. Canada (Minister of Health), 2008 FCA 108, 67 C.P.R. (4th)
23 at para. 53 [Pfizer Atorvastatin (FCA)]). In other words, does the
invention do what the patent promises it will do? The question to consider is
whether, at the date of filing, the patentee had sufficient information upon
which to base the promise. If not, the patentee must have had sufficient
information upon which to make a sound prediction of the promise.
[483]
At paragraph 70 of Wellcome AZT, above, the Supreme Court of Canada
articulated a three-part test that must be satisfied in order to establish that
a sound prediction has been made by the an inventor. The three elements of the
test are:
1.
there must be a factual basis for the prediction;
2.
the inventor must have an articulable and "sound" line of
reasoning from which the desired result can be inferred from the factual basis;
and
3.
there must be proper disclosure.
[484]
To be sound, a prediction does not need to amount to certainty, as it does
not exclude the risk that some compounds within the area claimed may, at some
later time, prove to be devoid of utility.
[485]
With these principles in mind, I turn to the '380 Patent and the evidence
before me.
(2) The '380 Patent
[486]
As of the relevant date of June 11, 1980 (the Canadian filing date), Merck
had made, but not tested, the compounds for which the process is claimed. In
other words, it was not relying on actual utility but on its prediction that
the four compounds would have utility.
[487]
As noted, sound prediction must be measured against the promise of the patent,
where one is explicitly expressed or may be implied. The promise of the '380
Patent is discussed in Section V of these Reasons. To review, I have concluded
the following:
1.
The '380 Patent does not promise that all micro-organisms within the
species Aspergillus terreus will produce the four compounds of claim 1
or the compounds identified in the other disputed claims.
2.
The patent does promise that the compounds produced by the fermentation
process identified in the patent are “useful as antihypercholesteremic agents
for the treatment of atherosclerosis, hyperlipemia and like diseases in
humans”.
[488]
I will measure the utility of the '380 Patent against these two promises.
(3) Lack of Utility
[489]
Apotex claims that it has presented empirical evidence of inutility of claims
1, 2 and 5. This assertion of inutility is founded on testing, by Drs. Sorenson
and Samson, that demonstrated that not all strains of micro-organisms within
the genus Aspergillus or – more narrowly – within the species Aspergillus
terreus are capable of producing the claimed compounds.
[490]
Dr. Sorensen was asked, by counsel for AFI, to investigate the ability of
different strains of Aspergillus terreus to produce lovastatin. He was
instructed to use fermentation conditions and media contemplated or
specifically taught in the '380 Patent. He designed experiments involving four
different strains of Aspergillus terreus. Two of those strains (referred
to as A18 and R99) were strains which had previously been shown to produce
Compound I. The other two strains (referred to as UAMH 7844 and UAMH 9313) were
obtained by Dr. Sorensen from the University of Alberta Microfungus
Collection and Herbarium (UAMH). In simple terms, Dr. Sorensen’s
results were as follows:
·
lovastatin (Compound1) was detected in the A18, R99 and UAMH 9313 extracts;
·
lovastatin was not detected in the UAMH 7844 extract; and
·
no quantifiable quantities of dihydrolovastatin (Compound II) were
detected in any of the four extracts.
[491]
Similar results were obtained by Dr. Samson, who also carried out testing
of certain micro-organisms at the request of counsel for AFI. Dr. Samson found
that:
·
A18, R99 and a strain identified as Merck ATCC 20542 produced lovastatin;
·
a strain identified as Aspergillus terreus IBT 20944 produced no
detectable quantities of lovastatin; and
·
a strain identified as Aspergillus alabamensis (which, according to
Dr. Samson, would have been classified as Aspergillus terreus in 1984) produced
no detectable quantities of lovastatin.
[492]
Witnesses presented by Merck did not dispute the core of these findings.
Dr. Alberts, one of the named inventors, acknowledged that not all Aspergillus
terreus strains tested by Merck produced lovastatin. Dr. Lasure agreed that
some strains of Aspergillus terreus would not produce lovastatin. In
Apotex’s view, the result must be that claims 1, 2 and 5, and each of the
dependent product claims 13 to 15, are invalid since they include embodiments
that will not achieve the promised result.
[493]
The problem with this argument is that Apotex relies on a construction and
promise of the '380 Patent with which I do not agree. During final argument,
Apotex conceded that their assertion is established assuming that “my friend’s
construction is not adopted”.
[494]
For the reasons expressed in my analysis of proper claims construction, I
have concluded that: (a) the claims only include micro-organisms that fall
within the species Aspergillus terreus; (b) the '380 Patent does not
promise that all micro-organisms within the species Aspergillus terreus
will produce lovastatin; and, (c) none of claims 1, 2 and 5 requires that all
four (or two, where applicable) compounds be produced from each fermentation.
Thus, it is irrelevant that that Drs. Sorenson and Samson found strains of
Aspergillus terreus which were incapable of producing lovastatin and
that other strains could only produce lovastatin (Compound I).
[495]
Apotex has failed to meet its burden to demonstrate that "[t]here is
evidence of lack of utility in respect of some of the area covered" (Wellcome
AZT, above, at para. 56).
(4) Sound Prediction
[496]
As noted above, Apotex may satisfy its burden of showing a lack of utility
even where it cannot demonstrate inutility. Its burden can be met by
demonstrating, on a balance of probabilities, that the prediction of the
inventors was not sound. I turn now to a consideration of whether the three
elements of the test for sound prediction, as set out in Wellcome AZT,
above, have been met.
[497]
In determining that the requirements for sound prediction had been met,
the Court in Wellcome AZT found that the factual basis for the sound
prediction of a new use compound rested upon the in vitro test results
of AZT against the HIV in a human cell line along with Glaxo's data on AZT,
including animal tests (above, para. 72). The line of reasoning was found to be
Glaxo's knowledge of the mechanism for the reproduction of a retrovirus.
(a) The Factual Basis
[498]
The question of sound prediction is one of fact (Wellcome AZT,
above, at para. 71). The inventors must be able to show that, at the relevant
time, they were in possession of a factual basis upon which they could
articulate the desired result. The perspective being examined at this stage is
a subjective one. The knowledge, activities and endeavours of the inventors
themselves must be considered.
[499]
In this case, Merck’s key witness was Mr. Alfred W. Alberts, one of the
named inventors of the '380 Patent. In credible testimony, Mr. Alberts told the
“story” of the invention that became the '380 Patent.
[500]
The '380 Patent story began in 1975, when Mr. Alberts arrived at Merck and
began working in the area that was called “basic research”. He established a
new department within that domain, a department that was called “biochemical
regulation”. The mandate of the department was to take a rational approach to
the development of new drugs. Mr. Alberts described the approach as follows:
The approach that I
was brought in to do was to go and do a – go back to the beginning, find the --
to break down the system, find the key targets for the disease, and then start
from there with discovering -- hopefully discovering compounds that affect the
disease process by working at the very simple, basic level, and then moving up
from there to animal studies.
[501]
One area of research for the department was cholesterol biosynthesis.
According to Mr. Alberts, it was well known, by 1975, that cholesterol was
intimately involved with the atherosclerotic process. The pathway of
cholesterol biosynthesis was understood in 1975 and was described by Mr.
Alberts as follows:
The pathway of
cholesterol biosynthesis is very complex. This is just a brief summary
highlighting the salient features of the process.
It starts with a
simple two carbon compound known as acetate and which is basically the salt of
vinegar.
It's activated to a
compound known as acetyl coenzyme A/acetyl CoA.
Then in a series of
steps, three acetyl CoA units are joined together to form the compound known as
hydroxymethylglutaryl coenzyme A/HMG-CoA, which is converted by an enzyme known
as HMG CoA reductase to mevalonic acid. And I will refer -- sometimes refer to
it as mevalonic acid or the salt form, which is mevalonate. And this six carbon
compound in a series of condensations ends up as the 30 carbon compound,
squalene, which is then modified into the 27 carbon sterol lipid cholesterol.
[502]
The Merck scientists were the first to isolate mevalonic acid – described
by Mr. Alberts as “the potential missing link in the cholesterol pathway”.
[503]
Based on their knowledge of this pathway, the Merck scientists were
looking for compounds that could break this chain. One part of the biosynthesis
that was of interest was the hydroxy-methylglutaryl-coenzyme A reductase (known
as HMG-CoA reductase) stage. The scientists understood that, if a compound
could inhibit or blocked the synthesis of cholesterol at the HMG‑CoA
reductase stage, there would be no conversion of: (a) HMG CoA reductase to
mevalonic acid; (b) mevalonic acid to squalene; and, (c) squalene to
cholesterol.
[504]
The Merck scientists first became aware of compounds that inhibited
HMG-CoA reductase in early 1976, when:
We received at Merck a
correspondence from a representative in Japan who came across a newly issued,
newly published patent in Japan describing an inhibitor of HMG-CoA reductase known as ML‑236B
and also became known as Compactin.
[505]
A sample of compactin was received from Sankyo Company in Japan. In
addition, Dr. Akira Endo of the Sankyo Company visited Merck on August 26,
1977. According to notes of the meeting, Dr. Endo presented data with respect
to ML-236B (compactin). It was stated that the compound “inhibits de novo
cholesterol biosynthesis and reduces serum cholesterol when administered orally
[in rats].” This compound operated as an HMG-CoA reductase inhibitor.
[506]
The goal was clearly to develop a compound that would be as good as or
better than compactin. Beginning in January 1978, the Merck scientists introduced
a new in vitro assay, the HMG‑CoA reductase assay. This allowed
the measurement of the inhibition of HMG-CoA reductase by any tested micro-organism.
ML-236B (compactin) was the benchmark against which the activity of organisms
from the Merck chemical library were measured. Between January 1978 and
November 1978, none of a large number of tested micro-organisms met the goal.
[507]
As reflected in the laboratory notebooks and in Mr. Alberts’s testimony,
November 7, 1978 was a turning point. In the first week of November 1978, Mr.
Alberts’s group received samples 18 and 19 (F 4683 and F 4684), both of which
demonstrated inhibitory activity in the HMG-CoA reductase assay.
[508]
From that point, the most important steps can be summarized as follows.
·
On November 27, 1978, Mr. Alberts strongly recommended that Merck further
pursue “the isolation and characterization of the inhibitory component in F
4683 for use as a potential hypocholesterolemic agent”.
·
In December 1978, another sample – F 4797 – was found to have inhibitory
activity that was tenfold higher than the first culture, F 4683.
·
On February 12, 1979, the structure of lovastatin, the lactone
(L-154,803), was recorded by Dr. Albers-Schonberg; the structure was similar to
compactin.
·
On February 12, 1979, Dr. Otto Hensens identified and recorded the
structure of the open dihydroxy acid form of lovastatin.
·
On February 13, 1979, Ms. Chen assayed the two samples used to identify
the structures above and confirmed that they were inhibitory.
·
On February 16, 1979, Mr. Alberts signed the Merck Confidential Memorandum
of Invention (referred to below).
·
On August 1, 1979, the structure of the natural dihydro (Compound II of
the '380 Patent) was identified and recorded by Dr. Otto Hensens, having been
found active in the HMG-CoA reductase assay on July 31, 1979.
·
On August 2, 1979, a hydrolyzed version of the natural dihydro, being the
open hydroxyl acid (Compound IV of the '380 Patent), was assayed and also found
to be active in the HMG-CoA reductase assay.
[509]
On February 16, 1979, Dr. Alberts completed a “Confidential Memorandum of
Invention”, on behalf of the inventors. This Memorandum summarizes the work of
the inventors. The structure of Compound I, a “homolog” of ML-236B (compactin),
“produced by an Aspergillus” is described. The inventors explicitly note
the utility or proposed use of the invention as “hypocholesteremic,
antifungal”. As of the date of the Memorandum, the inventors had found that the
compound had “in vitro potency similar to ML-236B as inhibiting HMG-CoA
reductase.”
[510]
Apotex’s main criticism of Merck’s work relates to what was not done by
the Merck scientists. Apotex submits that, because there was no testing of the
compounds on humans before the Canadian filing date, Merck was missing a
critical piece of factual information. In Apotex’s view, without this
information, the inventors did not have an adequate factual basis for the
prediction that the compounds would be “useful as antihypercholesteremic agents
for the treatment of atherosclerosis, hyperlipemia and like diseases in humans.”
However, when cross-examined on this issue, Mr. Alberts was clear as to how
Merck reached its prediction, even though human trials had not taken place by
the Canadian filing date.
Q. Based on the
test results that you had obtained, the in vivo animal test results you'd
obtained, would you not agree that you could not reliably predict that
Lovastatin was going to be an effective treatment for hypercholesteremia in
humans?
A. The only way
I could answer that is with any drug before it's been tested in humans, whether
it's Lovastatin, whether it's an antibiotic, no matter what it is, you can not
reliably predict it's going to work in humans until you put it into the humans,
and that is irrespective of the drug.
Q. In terms of
these particular models, these animal results, is it not fair to say that these
animal models, as a predictor of activity in humans, could not be relied upon
to make a sound an assessment of its potential effectiveness?
A. There was
enough -- let me go to one animal model that was a good predictor there, and
that's the dog, because the dog responds very nicely to the other cholesterol
lowering drug, cholestyramine; in fact, it's one of the few models that
responds to cholestyramine. Humans respond to cholestyramine. Rats do not
respond to cholestyramine. Dogs do. Humans do. So there was a reasonable
assumption that a drug that did not lower cholesterol in the rats would
conceivably work in humans and based on the biology, based on the biochemistry
of the system, it was a reasonable prediction that it would work and based on
our knowledge of Compactin. So we had a whole body of evidence that suggested
it would work.
[511]
I agree with Mr. Alberts; the inventors had a whole body of evidence that
suggested that the compounds would work in humans. The inventors had an
adequate factual basis for their predictions.
(b) Line of Reasoning
[512]
I next consider whether Apotex has shown that there was no articulable
line of reasoning from which the desired results could be inferred from the
factual basis. The question is: given the experimentation and laboratory
results that formed the factual basis together with information drawn from the
prior art, could Merck reasonably infer that the compounds would meet the
promise of being “useful as antihypercholesteremic agents for the treatment of
atherosclerosis, hyperlipemia and like diseases in humans”?
[513]
There was little expert evidence produced by either side that speaks
directly to the question of sound prediction or the line of reasoning. Some
assistance was provided by Dr. Gotto, who described his own observations and
the findings of scientists in the mid to late 1970s which demonstrated the link
between high cholesterol and atherosclerosis.
[514]
It was also known at the relevant time that HMG-CoA was the enzyme in the
liver responsible for making cholesterol.
[515]
Obviously, an essential element in the chain is an understanding of
cholesterol biosynthesis. No expert presented an alternative to Mr. Alberts’s description
of the biosynthesis pathway. It follows, from an understanding of cholesterol
biosynthesis, that a compound that can prevent the completion of this pathway –
at any stage – will have a good chance of preventing the formation of
cholesterol. Thus, it would have been part of the line of reasoning for the
development of all statins, such as lovastatin, that a compound that could
inhibit HMG-CoA in vivo could be predicted to lower cholesterol in
humans.
[516]
The next step in the chain of reasoning is the key element – that is,
compactin. The Merck inventors knew about the behaviour of compactin. Compactin
works on the cholesterol biosynthesis at the HMG CoA reductase stage; it
inhibits or prevents the enzyme from producing mevalonic acid. From the disclosure
contained in the Endo Patent, the Merck scientists had knowledge of the in
vivo activity of compactin.
[517]
The inventors also knew that the structure of the compounds that they had
developed from fermentation of Aspergillus terreus was similar to the
structure of compactin. It was not unreasonable for the inventors to predict
that a compound with a similar structure to compactin would have similar inhibitory
properties. Strengthening this prediction, the Merck scientists also had their
own in vitro testing data that demonstrated activity. All of this
information was available to the inventors of the '380 Patent as of February
12, 1979. This reasoning was confirmed by Dr. Gotto during cross‑examination:
Q. Let me ask
you this. If in 1979 a compound had been found that could inhibit HMG-CoA
reductase in a cell culture in vitro, would one be able to know whether or not
that compound would be effective in treating atherosclerosis, hyperlipidemia or
those kinds of diseases in humans?
A. Based on the
knowledge that one had about Compactin, yes.
[518]
In my view, an articulable line of reasoning has been demonstrated. In
other words, by February 12, 1979, the inventors of the '380 Patent could
soundly predict that the compounds of the invention would provide treatment for
hypercholesteremia in humans based on the known in vivo activity of the
closely-related compound, compactin, and Merck’s own in vitro data. In
other words, there was an articulable line of reasoning from which the desired
result – lowering of cholesterol in humans – could be inferred from the factual
basis. This was over a year in advance of the Canadian filing date of June 11,
1980.
[519]
Moving forward from February 1979, the Merck scientists were able to add
even more information to support the articulable line of reasoning. By July
1979, Merck had supplementary in vivo data that confirmed cholesterol
synthesis inhibition in rats. Well in advance of the Canadian filing date,
Merck had obtained positive results in dogs. This information, while not
essential to the sound prediction as of the Canadian filing date, certainly
provides additional support for the prediction.
(c) Disclosure
[520]
The final element of sound prediction is “disclosure”. The question is: in
the '380 Patent, has Merck provided disclosure of the factual basis and the
line of reasoning? I believe that it has.
[521]
In Eli Lilly Canada Inc. v. Apotex Inc., 2009 FCA
97, 78 C.P.R. (4th) 388 at para. 18, leave to appeal to SCC refused,
[2009] S.C.C.A. No. 219 (QL), 401 N.R. 400 (note)) [Eli Lilly Raloxifene
(FCA)], the Court of Appeal (referring to Wellcome AZT, above)
stated that, “where the claimed invention had not yet actually been reduced to
practice, the patent must provide a disclosure such that a person skilled in
the art, given that disclosure, could have as the inventors did, soundly
predicted that the invention would work once reduced to practice.”
[522]
Apotex’s argument of inadequate disclosure involves two discrete areas: the
first is the adequacy of disclosure of factual underpinnings of the invention,
and the second is the adequacy of disclosure of methods for producing the
claimed compounds from strains of Aspergillus.
[523]
With respect to the first argument, Apotex asserts that many of the facts
that were stated by Dr. Alberts to form the basis of the inventors’ prediction
of utility were not disclosed in the patent. Specifically, Apotex submits as
follows:
However, the only data
disclosed in the ’380 Patent that could form a “factual basis” for the
predicted utility of the compounds as anti-hypercholesteremic agents are the
tests reported in the examples. None of these tests evaluated the capacity of
a test compound to lower serum cholesterol levels in mammals or humans. None
of the compounds is shown to have been tested in an animal model to determine
whether it lowered serum cholesterol. There is no reference or explanation of
the rate-limiting role of the HMG-CoA reductase enzyme. There is no disclosure
of the knowledge that Merck acquired from its work with compactin, and no
disclosure of any relationship between the compounds of the invention and
cholestyramine, or why (and how) the properties of cholestyramine would inform
a prediction of utility. There is also no data about the toxicology,
pharmacokinetics or bioavailability of the compounds that would enable the
skilled addressee to predict that the compounds could be effectively
administered and tolerated by humans over the identified range of dosage
strengths. Accordingly, virtually all of the “facts” Dr. Alberts stated formed
the basis of the inventors’ prediction of utility are not disclosed.
[524]
The first problem with Apotex’s argument is that it is based on a
requirement that the patent disclose data to support the promise. The question
of whether or not a patentee has obtained enough data to substantiate its
invention is an irrelevant consideration with respect to the application of
subsection 27(3) of the Act. The Court is concerned with the sufficiency
of the disclosure, not the sufficiency of the data underlying the invention (Pfizer
Atorvastatin (FCA), above, at para. 56).
[525]
Apotex also refers to the animal testing carried out by Merck in 1979.
Reference is made to the testimony of Mr. Alberts where he agreed that the
testing of dogs led Merck to believe that lovastatin would result in
cholesterol reduction in humans. The '380 Patent does not disclose this
testing. Thus, Apotex submits that the failure to disclose the existence and
results of such tests establishes that there is insufficient information
disclosed in the '380 Patent to justify the prediction. I do not agree with
either Apotex’s characterization of Mr. Alberts’s testimony or the inference
that they draw.
[526]
Apotex relies on Eli Lilly Raloxifene (FC), above. In that case,
Justice Roger Hughes, the trial judge, found, as a matter of fact, that Merck
had placed reliance on a paper published before the Canadian filing date (the Hong Kong study).
Justice Hughes concluded that the requirement for disclosure had not been met;
the Court of Appeal agreed in Eli Lilly Raloxifene FCA, above, at paragraph 17.
As noted by the Court of Appeal in Eli Lilly Raloxifene FCA, at paragraph
15, “As the prediction was made sound by the Hong Kong study, this study had to
be disclosed.” In my view, the situation before me is distinguishable.
[527]
I agree that the cross-examination of Mr. Alberts resulted in a list of
facts and information that the inventors of the '380 Patent knew as of the
Canadian filing date. One of the areas of interest relates to the dog studies
carried out in 1979. Apotex pounces on this information as something that ought
to have been disclosed in the '380 Patent in order to justify the sound
prediction. However, I do not understand the jurisprudence to teach that the
patent specification must disclose absolutely everything that that the inventor
knew up to the relevant date. In Eli Lilly Raloxifene (FC), above,
without disclosure of the Hong Kong study, a skilled person would not have had sufficient
information to understand the justification for the prediction. We must examine
the specification to determine whether, with the information disclosed (even if
there was more information available and undisclosed), a skilled person could
have soundly predicted that the invention would work once reduced to practice.
[528]
In the case of the '380 Patent, the question is whether sufficient
information was disclosed to allow the skilled person to soundly predict that
the compounds of the invention would be “useful as antihypercholesteremic
agents for the treatment of atherosclerosis, hyperlipemia and like diseases in
humans”.
[529]
What was disclosed in the '380 Patent? The '380 Patent contains the
following disclosures:
·
the association between atherosclerosis and high cholesterol (p. 2, lines
11-15);
·
the utility of inhibiting cholesterol biosynthesis (p. 2, lines 14-16);
·
prior art consisting of US patents for compactin, a fermentation product
obtained from the genus Penicillium, which compound was found to be an
inhibitor, in vivo, of the biosynthesis of cholesterol (p. 2, lines
17-26);
·
the discovery by the inventors of the '380 Patent that compounds produced
from Aspergillus are more potent inhibitors of cholesterol synthesis in
vivo than compactin (p. 2, lines 28-33; p. 3, lines 1-5);
·
the fact that HMG-CoA reductase inhibition is the relevant means of
inhibiting cholesterol biosynthesis (p. 14, lines 30-35);
·
the structure of lovastatin and the other compounds (p. 9, 10, 11 and 12);
·
in vitro HMG-CoA reductase inhibition by
lovastatin, (p. 14, line 30; p. 15, line 6; p. 24, lines 7-8; p. 25, lines
10-12; p. 42, lines 1-13); and
·
in vivo inhibition of cholesterol synthesis by Compound II (p. 42,
lines 15-25).
[530]
Taken as a whole, I am not persuaded that there was inadequate disclosure.
A skilled person would conclude that the '380 Patent sufficiently discloses the
factual basis and the line of reasoning to soundly predict that the claimed
compounds would be useful in the treatment of high cholesterol. In other words,
as required by the jurisprudence, the '380 Patent discloses the factual basis
and line of reasoning for its promise. The prediction was made sound by the
information disclosed; the disclosure of other information within the knowledge
of the inventors was not essential to the prediction.
[531]
The second of Apotex’s submissions is that the '380 Patent fails to
disclose the methods for determining which strains of the genus Aspergillus
will produce the desired compounds. As discussed in the section of these
reasons on claims construction, I have concluded that the patent only claims
compounds produced from Aspergillus terreus and, further, that it does
not promise that lovastatin can be produced from all strains of Aspergillus
terreus. Apotex argues that, even on this narrower construction and
promise, Merck was required to disclose the methods for identifying producing
strains. Apotex argues that the specification does not disclose this
information and that to find the producing strains would require excessive and
inventive experimentation by the skilled person.
[532]
The courts have recognized that “routine trials and experiments not
amounting to new inventions might be required to put [an invention] into
practice” (Proctor & Gamble Co. v. Bristol-Myers Ltd.(1978), 39
C.P.R. (2d) 145 at para. 51, [1978] F.C.J. No. 812 (QL) (F.C.T.D.); see also, Mobil
Oil Corp. v. Hercules Canada Inc. (1995), 63 C.P.R. (3d) 473, [1995] F.C.J.
No. 1243 (QL) (F.C.A.)); Aventis Pharma Inc. v. Apotex Inc. 2005 FC 1283,
43 C.P.R. (4th) 161 at para. 207). The evidence before me does not support
Apotex’s assertion that inventive experimentation would be required to find
producing strains of Aspergillus terreus. I have already discussed this
issue (see paragraphs 57 to 130) under claims construction. To repeat, I am
satisfied that the skilled person could use well-known techniques to rapidly
screen a large number of isolates of strains of Aspergillus terreus to
determine which strains are producing. Moreover, since, on my construction, the
claims are limited to strains of the species Aspergillus terreus, there
are manageable boundaries on the testing that would be required.
D. First
Inventorship/Missed Conflict
(1) Introduction
[533]
Apotex submits that the Merck inventors were not the first inventors of
the compound lovastatin as claimed in the '380 Patent and that, therefore, the
'380 Patent should be invalidated on the basis that the “invention” of the '380
Patent was known or used as of the date of filing the patent application. In so
arguing, Apotex recognizes that it must overcome the requirements of the Patent
Act.
[534]
The patent application that resulted in the issuance of the '380 Patent
was patent application No. 353, 777 filed with the Patent Office on June 11,
1980 (the Monaghan Application). Apotex submits that the Monaghan Application
disclosed the invention of lovastatin and ought to have been placed into
conflict with patent application No. 345, 983 (the Endo Application), which was
filed with the Patent Office on February 19, 1980. The Endo Application
ultimately resulted in the issuance of the ‘794 Patent.
(2) Legal Principles
[535]
Prior to 1989, the overall scheme under the Patent Act was
one of first to invent. By contrast, the scheme under the Patent Act currently
in force can be described as first to file. The notion of first inventorship is
embodied in s. 27(1) of the Act:
|
27. (1) Subject to this section, any inventor or legal
representative of an inventor of an invention that was
(a)
not known or used by any other person before he invented
it,
(b)
not described in any patent or in any publication printed
in Canada or in any other country more than two years before presentation of
the petition hereunder mentioned, and
(c)
not in public use or on sale in Canada for more than two
years prior to his application in Canada,
may, on presentation to the Commissioner of a petition
setting out the facts, in this Act termed the filing of the application, and
on compliance with all other requirements of this Act, obtain a patent
granting to him an exclusive property in the invention.
|
Sous réserve des autres dispositions du présent article,
l'auteur de toute invention ou le représentant légal de l'auteur d'une
invention peut, sur présentation au commissaire d'une pétition exposant les
faits, appelée dans la présente loi "le dépôt de la demande", et en
se conformant à toutes les autres prescriptions de la présente loi, obtenir
un brevet qui lui accorde l'exclusive propriété d'une invention qui n'était
pas :
a)
connue ou utilisée par une autre personne avant que
lui-même l'ait faite;
b)
décrite dans un brevet ou dans une publication imprimée
au Canada ou dans tout autre pays plus de deux ans avant la présentation de
la pétition ci-après mentionnée;
c)
en usage public ou en vente au Canada plus de deux ans
avant le dépôt de sa demande au Canada.
|
[536]
Recognizing that more than one person might claim inventorship to similar
or overlapping subject matters, Parliament provided means for identifying and
resolving such a conflict. To begin, s. 43(1) of the Act defines when a
conflict exists:
|
Conflict between two or more pending applications exists
(a)
when each of them contains one or more claims defining
substantially the same invention; or
(b)
when one or more claims of one application describe the
invention disclosed in one of the other applications
|
Se produit un conflit entre deux ou plusieurs demandes
pendantes dans les cas suivants:
a)
chacune d'elles contient une ou plusieurs revendications
qui définissent substantiellement la même invention;
b)
une ou plusieurs revendications d'une même demande
décrivent l'invention divulguée dans l'autre ou les autres demandes
|
The balance of s. 43 sets out the
procedures for declaring and dealing with a conflict.
[537]
While s. 27(1) gives the right to a patent to the first inventor, the Act
also contemplates that legal proceedings may be brought with respect to the
validity of patents (see the Patent Act, starting at s. 53). In
particular, s. 59 of the Act permits a defendant (such as Apotex) in a
patent infringement action to plead "any fact or default which by this Act
or by law renders the patent void." Under s. 60(1) of the Act, a
patent or any claim in a patent may be "declared invalid or void by the
Federal Court ... at the instance of any interested person."
[538]
However, when the validity of a patent is being challenged on the question
of inventorship, s. 61(1) is a limiting or qualifying provision. In the case
before me, s. 61(1)(b) is relevant:
|
No patent or claim in a patent shall be declared invalid
or void on the ground that, before the invention therein defined was made by
the inventor by whom the patent was applied for, it had already been known or
used by some other person, unless it is established that
(b)
that other person had, before the issue of the patent,
made an application for patent in Canada on which conflict proceedings should
have been directed;
|
Aucun brevet ou aucune revendication dans un brevet ne
peut être déclaré invalide ou nul pour la raison que l'invention qui y est
décrite était déjà connue ou exploitée par une autre personne avant d'être
faite par l'inventeur qui en a demandé le brevet, à moins qu'il ne soit
établi que, selon le cas :
b)
cette autre personne avait, avant la délivrance du brevet,
fait une demande pour obtenir au Canada un brevet qui aurait dû donner lieu à
des procédures en cas de conflit;
|
[539]
As stated in s. 61(1)(b), no patent will be declared invalid on the
grounds of prior inventorship by some other person unless the challenging
party can establish that the other person had, before the issue of the
patentee's patent, made an application for a patent in Canada on
which conflict proceedings should have been directed. Stated in other
words, a party may only successfully raise inventorship as an issue if: (a) the
invention in the patent or claim had already "been known or used by some
other person"; (b) the other person made a patent application for this
prior invention in Canada; or (c) "conflict proceedings should have been
directed". The interpretation of s. 61(1)(b) was the subject of discussion
in the case of Laboratoires Servier v. Apotex Inc., 2008 FC 825,
67 C.P.R. (4th) 241 [Servier FC] .
[540]
Thus, the threshold question to be answered is whether, on the facts
before me, there was a “missed conflict”. In other words, should
conflict proceedings have been directed between the Endo Application and the
Monaghan Application? If there was no missed conflict, Apotex is precluded, by
s. 61(1)(b), from challenging the validity of the '380 Patent on the grounds
of prior inventorship.
(3) Was there a missed conflict
[541]
The Endo Application was filed on February 19, 1980; the Endo Patent was issued
on August 17, 1982. The Monaghan Application was filed on June 11, 1980;
the ‘380 Patent was issued on January 31, 1984. Conflict could have been
declared only during the co-pendency of the two applications; that is, between
June 11, 1980 and August 17, 1982.
[542]
A review of the case or file history for the '380 Patent provides us with
the sequence of events leading to the issuance of the patent.
[543]
As filed, claims 1 to 7 of the Monaghan Application were claims to
processes for producing certain compounds. However, claim 8 was for Compounds I
and II alone and claim 9 was for Compounds III and IV alone. Claims 10 to 19
were claims to salts, esters and compounds all of which were dependent on
claims 8 or 9. While claims 1 to 7, as filed, were product-by-process claims, claims
8 to 19 did not contain process restrictions; they were what is referred to as per
se claims. The problem with per se claims is that they were not
permissible under the Patent Act, as it
existed at the relevant time.
Specifically, s. 41(1) (later, s. 39(1)) of the Patent Act, as it stood
in 1982) stated that:
|
In the case of inventions relating to naturally occurring
substances prepared or produced by, or significantly derived from,
microbiological processes and intended for food or medicine, the
specification shall not include claims for the resulting food or medicine
itself, except when prepared or produced by or significantly derived form the
methods or processes of manufacture particularly described and claimed.
[Emphasis added.]
|
Lorsqu'il s'agit d'inventions couvrant des substances que l'on trouve dans la nature, préparées ou produites, totalement ou pour une part notable, selon des procédés microbiologiques et destinées à l'alimentation ou à la médication, aucune revendication pour l'aliment ou le médicament ne doit être faite dans le mémoire descriptif,
sauf pour celui ainsi préparé ou produit selon les modes du procédé de
fabrication décrits en détail et revendiqués.
[Non souligné dans l’original.]
|
In simple terms, the inventors of the
Monaghan Application could never have received a patent that included claims 8
to 19, as originally filed.
[544]
In a letter from the Patent Office dated November 10, 1982 (Office
Action), counsel for the Monaghan Application was advised of the problem:
The claims of this
application are governed by Section 41(1) of the Patent Act. In the case of a
substance intended for food or medicine and prepared by a chemical process, an
application must have a patentable process claim, and any product claim must be
in process-dependent form and of the same scope as the process claim. Amendment
or cancellation of claims 8-19 is required.
[545]
The response to the Office Action was received by the Patent Office on
February 9, 1983. Amended claims 1 through 20 were substituted for the
non-allowable claims 8 to 19. Evidently, the Commissioner of Patents agreed
with the submission that the Monaghan Application now contained “claims
conforming with the requirements of section 41(1)”; the '380 Patent was issued on
January 31, 1984.
[546]
Each of Apotex and Merck put forward an expert to speak to the practices
of the Patent Office at the relevant times. Both Mr. Robert Barrigar (put
forward by Merck) and Mr. Robert Hirons (put forward by Apotex) have been
registered patent agents in Canada for a long time. Both have extensive knowledge of Canadian
patent prosecution and practice between 1980 and 1982, and are knowledgeable
about practices of the Commissioner of Patents at that time. I accepted the
qualifications of each to speak to these matters.
[547]
Both experts agreed that a declaration of conflict would not have been
made between two pending applications when each application contained
process-dependent claims for the same compound made using different processes. Accordingly,
the question of missed conflict could only have arisen between claims 8 to 19,
as originally filed, of the Monaghan Application and some or all of the claims
of the Endo Application.
[548]
Mr. Hirons opined that the requirements of s. 43(1)(b) of the Patent Act
were satisfied and a conflict, as defined in that provision, existed between
the two applications (Hirons Expert Report, Exhibit 117, para. 13). Mr. Hirons
pointed out that, during the period of co-pendency, the compound claims in the
Monaghan Application were not restricted by process. Thus, in his view, a conflict
between the two applications necessarily existed. In other words, one or more
of claims 8 to 19 of the Monaghan Application described the invention disclosed
in the Endo Application and, as a result, the Commissioner should have
commenced conflict proceedings.
[549]
Frankly, this is, to a large degree, a legal opinion and not one for which
I need the assistance of an expert. That being said, I do not disagree with Mr.
Hirons’s conclusion that a conflict, as defined in s. 43(1)(b), existed as
between the two applications. Moreover, Mr. Hirons’s description of how
conflict proceedings were conducted, once directed, is not inaccurate.
[550]
However, Mr. Hirons fails to answer the question of whether “conflict
proceedings should have been directed”. He assumes that the existence of
a conflict automatically requires the Commissioner to direct conflict
proceedings. The mere existence of a conflict at the application stage does
not, in my view, automatically mean that conflict proceedings should have been
directed.
[551]
In responding to this question, I refer first to the procedures described
in s. 43, in respect of the legal requirements of the Patent Act.
Sections 43(2) to 43(4) deal with procedures to be followed before a
conflict is declared. I accept that, when a s. 43(1) conflict exists, s. 43(2)
sets out a mandatory procedure to be followed.
|
Procedure to be followed before conflict is declared
(2) When the Commissioner has before him two or more applications
referred to in subsection (1), he shall
(a) notify each of the application of the apparent conflict and
transmit to each of them a copy of the conflicting claims, together with a
copy of this section; and
(b) give to each applicant the opportunity of inserting the same
or similar claims in his application within a specified time.
|
Procédure à suivre avant déclaration de conflit
(2) Lorsque le commissaire a devant lui deux ou
plusieurs de ces demandes, il doit :
a) notifier à chacun des demandeurs le conflit
apparent, et transmettre à chacun d’’ eux une copie des revendications
concurrentes, ainsi qu’une copie du présent article;
b) procurer à chaque demandeur l’occasion d’insérer dans
sa demande les mêmes revendications ou des revendications similaires, dans un
délai spécifié.
|
[552]
In this case, assuming that claims 9 to 19 of the Monaghan Application, as
originally filed, were in conflict with some or all of the claims in the Endo Application,
the Commissioner was required to follow the procedure set out in s. 43(2) of
the Act. He did not do so in this case. However, these are steps to be
taken before the declaration of conflict to determine if there should be a formal
declaration. Thus, even if the Commissioner erred by not complying with the
mandatory provision, any such error would be of no moment if, at the end of the
day, there was no need to make a formal declaration of conflict. The next step
– the formal declaration – is set out in s. 43(5):
|
Formal declaration of conflict
(5) Where the subject matter of the claims described in
subsection (3) is found to be patentable and the conflicting claims are
retained in the applications, the Commissioner shall require each applicant
to file in the Patent Office, in a sealed envelope duly endorsed, within a
time specified by him, an affidavit of the record of invention, which
affidavit shall declare
(a)
the
date at which the idea of the invention described in the conflicting claims
was conceived;
(b)
the
date on which the first drawing of the invention was made;
(c)
the
date when and the mode in which the first written or oral disclosure of the
invention was made; and
(d)
the
dates and nature of the successive steps subsequently taken by the inventor
to develop and perfect the invitation from time to time up to the date of the
filing of the application for patent.
|
Déclaration formelle de conflit
(5) Si l’objet des revendications visées au
paragraphe (3) est reconnu brevetable et que
les revendications concurrentes sont maintenues dans les
demandes, le commissaire exige de chaque demandeur le dépôt, au
Bureau des brevets, dans une enveloppe scellée portant
une suscription régulière, dans un délai qu'il
spécifie, d'un affidavit du relevé de
l’invention. L'affidavit déclare :
a) la date d laquelle a été conçue l’idée
de l'invention décrite dans les revendications
concurrentes;
b) la date laquelle a été fait le premier dessin
de l’invention;
c) la date i laquelle a été faite 1a première
divulgation écrite ou orale de !'invention et la manière dont elle a été
faite;
d) les dates et la nature des expériences
successives que l’inventeur a pratiquées par la suite afro de développer et
mettre graduellement an point cette invention jusqu'à la date du dépôt de la
demande de brevet,
|
[553]
Of critical importance, this provision establishes that the conflict
proceedings described in s. 43(5) and the balance of s. 43 only apply
“where the subject-matter of the claims described in subsection (3) is found to
be patentable and the conflicting claims are retained in the applications.”
[554]
Relating the file history to the co-pendency period between the Endo
Application and the Monaghan Application, at no time between June 11, 1980 and
August 17, 1982 was the Monaghan Application in a patentable form. Quite simply
the potentially-conflicting per se claims of the Monaghan Application
were not patentable; claims 8 to 19 did not contain claims that met the
requirements of the Patent Act. The subject matter of claims 8 to 19 was
not patentable. In short, during the co-pendency period, there was no
obligation on the Commissioner to declare a conflict.
[555]
My understanding of s. 43 of the Patent Act is consistent with the
practices of the Patent Office. Mr. Barrigar, using his experience and
knowledge of processes in the Patent Office, provided his opinion on how, in
practical terms, the Patent Office would have dealt with the two patent
applications. This portion of his Expert Report and his oral testimony were
very helpful. In brief, the Patent Office would not, faced with unpatentable
claims, have initiated conflict
proceedings. Rather, the practice of the
Patent Office was as described by Mr. Barrigar (Barrigar Expert Report, Exhibit
44, para. 18):
Based on my
experience, and consistent with the provisions of the "Old Act", it
was the uniform practice of the Patent Office in the period 1980 to and
including 1984 to implement conflict practice in the context of the Patent Act
as a whole. As discussed more fully below, this meant that only claims
satisfying the other requirements of the Act and applicable Patent Rules were
placed in conflict. It was the practice of the Patent Office to endeavour to
dispose of all other claiming issues before declaring any conflict, so that if
possible it would not be necessary to conduct conflict proceedings. Conflict
proceedings constituted a drain on Patent Office resources and delayed issue of
patents.
[556]
Thus, while the Commissioner may “technically” have been required to
comply with ss. 43(2) to 43(4) of the Act, his failure to do so is
without consequence where, as in this case, the requirements to pursue the
conflict in formal proceedings, as set out in s. 43(5), were not met. Even Mr.
Hirons finally agreed, on cross-examination that:
There are very good
policy reasons why the Commissioner of Patent would not want to launch or
commence conflict proceedings in respect of claims that everybody knows cannot
be issued in the form they are written.
[557]
Apotex asserts that the practice of resolving apparent conflicts, by
requiring the applicant to amend all or part of the specification to remove
objectionable claims, should have no bearing on the right to invoke s. 61 of
the Patent Act. I disagree. Where a patent application contains claims
that can never result in a patent, there is no conflict to declare.
[558]
In summary on this issue, I am satisfied that, on the facts of this case,
there was no requirement to direct conflict proceedings between the Endo
Application and the Monaghan Application. There was no missed conflict and
Apotex is precluded, by s. 61(1)(b) of the Act, from challenging the
validity of the '380 Patent on the grounds of prior inventorship.
(4) Did the Endo application disclose the invention of the '380
Patent?
[559]
If I am incorrect in my conclusion that Apotex is precluded from challenging
the validity on the grounds of prior inventorship, I turn to the question of
whether the Endo Application and Patent disclose the same invention as that of
the '380 Patent.
[560]
The Endo Application and Patent clearly refer to a substance called “Monacolin
K”. As set out in claim 1, the Endo Patent claims:
A process for
preparing Monacolin K, which process comprises cultivating a Monacolin
K-producing micro-organism of the genus Monascus in a culture medium
therefore.
[561]
It is accepted that Monacolin K is lovastatin. However, as discussed earlier
in these Reasons, an essential feature of the claims of the '380 Patent is that
the lovastatin be made using a strain of the species Aspergillus terreus.
The invention is lovastatin when made by fermentation of Aspergillus terreus
and does not include lovastatin made with other micro-organisms. Monascus
is a different genus from Aspergillus. Accordingly, the two applications
that included allowable claims did not contain one or more claims defining
substantially the same invention. Moreover, neither application described the
invention disclosed in the other application.
[562]
Thus, even if I accept that there was a missed conflict, the invention of
the '380 Patent is not the same as that of the Endo Application or Patent.
Apotex has not persuaded me that the invention of the '380 Patent was known or
used by Dr. Endo prior to the filing date of the Monaghan Application.
(5) Red Yeast Rice/Anticipation
[563]
Apotex submits that claims 13 to 15 of the '380 Patent are invalid
pursuant to the principles of “anticipation by prior use”. Briefly stated,
Apotex’s argument is that the compound lovastatin (also known as Monacolin K)
was known and used in the form of traditional Red Yeast Rice long before the
priority date of the '380 Patent or any earlier date of invention. Thus, Apotex
alleges that the presence of lovastatin in any Red Yeast Rice product before
the priority date of the '380 Patent was anticipatory of claims 13 to 15. In
this section, I will use the term “Red Yeast Rice” to refer to the products,
including rice, that are made with the substance known as red yeast.
(a) Principles
of Anticipation
[564]
The concept of anticipation arises from s. 27(1) of the Patent Act.
Subsection 27(1) permits any inventor to file an application for an invention
that was “not known or used by any other person before he invented it”. This
requirement is echoed in s. 61(1)(a), which allows the Court to invalidate any
patent or claim(s) if another person had “disclosed or used the invention in
such a manner that it had become available to the public”. In short, the Patent
Act requires that the subject matter of a claim must not have been disclosed
to the public before the claim date.
[565]
The guiding jurisprudence on the legal test of anticipation is found in
the Supreme Court of Canada decision in Apotex v. Sanofi-Synthelabo, 2008 SCC
61, [2008] 3
S.C.R. 265 [Sanofi-Synthelabo]. At paragraphs 23-27, the
Supreme Court teaches that the issue of whether an invention is anticipated by
the prior art requires that the Court have regard to two questions:
1.
Was the subject matter of the invention disclosed to the public by a
single disclosure?
2.
If there has been such a clear disclosure, is the working of the invention
enabled by that disclosure?
[566]
At the first step of the analysis, the Supreme Court provided the
following guidance (Sanofi‑Synthelabo, above, at para. 25):
When considering the
role of the person skilled in the art in respect of disclosure, the skilled
person is "taken to be trying to understand what the author of the
description [in the prior patent] meant" (para.32). At this stage, there
is no room for trial and error or experimentation by the skilled person. He is
simply reading the prior patent for the purposes of understanding it.
[567]
Once disclosure has been made, the question of enablement was described by
the Supreme Court (Sanofi-Synthelabo, above, at para 27):
Once the subject matter
of the invention is disclosed by the prior patent, the person skilled in the
art is assumed to be willing to make trial and error experiments to get it to
work. While trial and error experimentation is permitted at the enablement
stage, it is not at the disclosure stage. For purposes of enablement, the
question is no longer what the skilled person would think the disclosure of the
prior patent meant, but whether he or she would be able to work the invention.
[568]
In Abbott Laboratories v. Canada (Minister of Health), 2008 FC
1359, 337 F.T.R.
17 [Abbott Clarithromycin (FC)], aff'd 2009 FCA
94, 387 N.R.
347, Justice Hughes undertook a helpful survey of the law of
anticipation as it exists after Sanofi-Synthelabo, above. He summarized
the legal requirements for anticipation as follows (Abbott Clarithromycin
(FC), above, at para. 75):
For there to be
anticipation there must be both disclosure and enablement of the claimed
invention.
1.
The disclosure does not have to be an "exact description" of the
claimed invention. The disclosure must be sufficient so that when read by a
person skilled in the art willing to understand what is being said, it can be
understood without trial and error.
2.
If there is sufficient disclosure, what is disclosed must enable a person
skilled in the art to carry out what is disclosed. A certain amount of trial
and error experimentation of a kind normally expected may be carried out.
3.
The disclosure when carried out may be done without a person necessarily
recognizing what is present or what is happening.
4.
If the claimed invention is directed to a use different from that
previously disclosed and enabled then such claimed use is not anticipated.
However if the claimed use is the same as the previously disclosed and enabled
use, then there is anticipation.
5.
The Court is required to make its determinations as to disclosure and
enablement on the usual civil burden of balance and probabilities, and not to
any more exacting standard such as quasi-criminal.
6.
If a person carrying out the prior disclosure would infringe the claim
then the claim is anticipated.
[569]
The date for assessment of anticipation is June 15, 1979, the priority
date of the '380 Patent.
(b) Background on Red Yeast Rice
[570]
It is undisputed that Red Yeast Rice (or similar products containing red
yeast, such as red yeast bean curd) has been used produced and consumed in
Asian countries for hundreds of years. Dr. Scott Harding, an expert witness presented
by Apotex, described the uses of Red Yeast Rice in (mainly) Chinese culture as
follows (Harding Expert Report, Exhibit 115, para. 15):
[Red Yeast Rice] has
traditionally been used as a specialty food, as a dye or food pigment and as a
natural remedy for gastrointestinal infections and diseases of the blood.
[571]
According to Dr. Harding, more recently, Red Yeast Rice has been marketed
as a product that can lower lipid levels and reduce cardiovascular risk.
(c) Legal Consequences of lovastatin in Red Yeast
Rice
[572]
The expert testimony of Dr. Harding is to the effect that Red Yeast Rice
is not and was not produced from Aspergillus terreus. Dr. Harding opined
that traditional Red Yeast Rice was produced through fermentation of certain
strains of species from the genus Monascus.
[573]
Since Red Yeast Rice does not involve the use of Aspergillus terreus
in any way, it cannot have been an anticipatory disclosure of the process
claims of the '380 Patent. As discussed in the section of these reasons on
claims construction, claims 1 to 12 are process claims in which producing
strains of Aspergillus terreus are fermented to produce certain
compounds (including lovastatin as Compound I). Apotex does not assert the
argument of anticipation against claims 1 to 12; rather it limits it argument
to the product-by-process claims 13 to 15.
[574]
Claim 13 is a claim to any one of Compounds I to IV, when prepared by the
process defined in claim 1 or by an obvious chemical equivalent. Claim 14 is a
claim to one of Compound I or II, when prepared by the process defined in claim
2 or by an obvious chemical equivalent and claim 15 is a claim to one of
Compound III or IV, when prepared by the process defined in claim 5, or by
an obvious chemical equivalent. Earlier in these reasons, I construed claims 13
to 15 to require that, as an essential element, the product (be it Compound I,
II, III or IV) is produced by the process defined in the earlier claims.
Specifically, the compounds of claims 13 to 15 must be made from the
fermentation of a species of Aspergillus terreus.
[575]
For the moment, I will assume that lovastatin could be found in Red Yeast
Rice as of the priority date. I accept the opinion of Dr. Harding that Red
Yeast Rice is a result of the fermentation of species within the genus Monascus.
If the allegedly anticipatory product – lovastatin found in Red Yeast Rice – is
produced from something other than Aspergillus terreus, it cannot, in my
view, meet the legal test for anticipation. Stated in terms of the test, Red
Yeast Rice does not meet the first requirement that the prior invention must
disclose the subject matter of claims 13 to 15.
[576]
Apotex responds to this analysis by asserting that a product-by-process
claim is a claim to the product. In the words of Apotex:
Where the product was
previously known and a new process for making it has been discovered, the only
invention that can be claimed is the process because the product is not “new”.
Thus, a product-by-process claim can be anticipated by prior disclosure of the
product but not of the process.
Apotex relies on the Supreme Court
decision in Hoffmann-LaRoche Ltd. v. Canada
(Commissioner of Patents), [1955] S.C.R. 414, 23 C.P.R. 1 [Hoffmann-LaRoche 1955
cited to S.C.R.] for this submission.
[577]
I acknowledge that the Hoffmann-LaRoche 1955 decision does
appear to support Apotex’s position. Hoffmann-LaRoche 1955 involved an
application for a patent that claimed a new process for making a known
substance called aldehyde, as well as aldehyde when made by that process. The
Commissioner of Patents granted the claim for the new process for making
aldehyde, but not the claim for aldehyde made by that process. The inventor
appealed to the Exchequer Court, without success (Hoffmann-La Roche
Ltd. v. Canada (Commissioner of Patents)(1953), [1954] Ex.
C.R. 52, 19 C.P.R. 80), and then to the Supreme Court of Canada,
again without success. In his brief reasons for dismissing the appeal, Chief
Justice Cartwright, speaking on behalf of four of five of the justices, stated:
“There being nothing new about the product, the appellant is not entitled to
obtain a patent therefore even on the basis of a process dependent product
claim” (Hoffmann-LaRoche 1955, above, at p.415). Little analysis
is offered in the single page of Chief Justice Cartwright’s reasons.
[578]
I admit to having considerable difficulty in understanding how the
conclusion in Hoffmann-LaRoche 1955 can fit with the protection offered
by the Patent Act. It seems illogical to me that a process for making a
substance can be novel and thus patentable but that a claim for the product
when made by that process is automatically not patentable. I understand that
situations could exist that would invalidate the product-by-process claim. For
example (without limitation):
·
a product-by-process claim could be challenged on the basis that the
substance, when made by the described process, was anticipated or
obvious; or
·
an earlier patent or disclosure is to the product made by any means
(although, of course, this could not have happened with the '380 Patent since,
at that time, per se claims were not permitted).
However, where the substance is only
claimed when made by a particular process and where the making of the product
by that particular process is novel, the boundaries of the claim are well
delineated. I cannot see why the claim would automatically be invalidated
simply because someone else has claimed the same product made by a different
process.
[579]
Finally, I observe that the Supreme Court in Hoffmann-LaRoche 1955 implicitly
accepted the validity of the claims to “a new and useful process for
manufacture of an aldehyde”. There is, in my view, no principled reason why, if
the product-by-process claim is invalidated, the new process for making the
known substance is not also invalid.
[580]
There has been little jurisprudence in which the principle in Hoffmann-LaRoche
1955. However, what little there has been over the last 65 years does not,
it seems, directly contradict the Supreme Court decision. Hoffmann-LaRoche
1955 was recently considered by the Federal Court of Appeal in Abbott
Laboratories v. Canada (Minister of Health), 2007 FCA
153, 59 C.P.R. (4th) 30 (Abbott Clarithromycin (FCA)), where Justice
Sharlow observed at paragraph 15 that, “There is no
jurisprudence that casts any doubt on the correctness of the principle stated
in Hoffmann.”
[581]
The Court of Appeal also had occasion to examine the applicability of Hoffmann-LaRoche
1955 in Bayer AG v. Apotex Inc., 2001 FCA 263, 14 C.P.R. (4th)
263 [Bayer Ciproflaxin]. In that case, Apotex Inc. appealed from an order
prohibiting the Minister of National Health and Welfare from issuing to it an
NOC under the Regulations. Apotex alleged that Bayer's Canadian patents
for ciprofloxacin were invalid because Bayer had already applied in another
country for a patent on the same drug. The inventions were substantially the
same but used a different synthesis process. Apotex Inc. relied on Hoffmann-LaRoche
1955 for the proposition that the differences in the process for the making
of the drug were not “legally relevant” (Bayer Ciproflaxin, above, at
para. 15). In rejecting this argument, Justice Evans, speaking for the Court,
commented as follows (Bayer Ciproflaxin, above, at para. 16):
In our view, however,
that case is readily distinguishable from the case at bar. In particular, the
issue in Hoffmann-La Roche, supra, was whether the patent application satisfied
the requirement in paragraph 28(1)(b) of the Patent Act that a patent may only
be obtained for an invention to the extent that it contains an element of
novelty. This is not the question before us. What we must decide is whether the
inventions that are the subject of the Chilean and Canadian patents are the
same invention. And, as counsel for Bayer points out in his memorandum, Apotex
did not allege in any of its notices of allegations that the product by process
patent obtained in Canada for ciprofloxacin was invalid because ciprofloxacin
had already been invented, and that the use of the malonic ester synthesis
process to produce an intermediate could not therefore render the invention
novel. [Emphasis added.]
[582]
In summary, it appears that I must accept the holding of Hoffmann-LaRoche
1955. In contrast to the situation before Justice Evans in Bayer
Ciproflaxin, Apotex has clearly argued that the product-by-process claims
are invalid because the substance – lovastatin or Monacolin K – had already
been invented. Paraphrasing the words of Justice Sharlow in Abbott
Clarithromycin (FCA), above, for the purposes of applying the Hoffmann-LaRoche
1955 principle, lovastatin is "known" as of June 15, 1979, if a hypothetical
claim for its invention would fail on the ground of anticipation or lack of
novelty.
[583]
Apotex submits that the existence of Red Yeast Rice – a substance known
for centuries – would satisfy this test. The issue is whether the product
lovastatin, as described in the '380 Patent, however made, was anticipated by
Red Yeast Rice.
(d) Evidence of lovastatin in Red Yeast Rice prior to the
priority date
[584]
The first question is whether the evidence before me establishes that Red
Yeast Rice was a source of lovastatin before the priority date. If it does not,
then the claim of anticipation must fail. However, if I am persuaded that, on a
balance of probabilities, Red Yeast Rice, prepared by traditional methods,
contained lovastatin, I will continue on to consider whether there was both
disclosure and enablement.
[585]
In Dr. Harding’s opinion, there is ample scientific literature that
indicates that Red Yeast Rice and similar products “produced in solid phase
fermentation and using strains of Monascus that were traditionally
employed, have appreciable levels of Monacolin K [lovastatin] and are effective
in lowering blood cholesterol in vivo” (Harding Expert Report, Exhibit 115,
para. 52). Dr. Harding concludes that, “people have consumed Monacolin K for
centuries.”
[586]
Dr. Harding stated that the amount of Monacolin K in Red Yeast Rice will
vary depending on the fermentation conditions used. It follows (and Dr. Harding
did not disagree) that not all Red Yeast Rice contains Monacolin K.
[587]
In addition to his literature search, Dr. Harding tested two different
samples of Red Yeast Rice products that he bought from Asian specialty markets
in Winnipeg and Toronto. His conclusion was that both samples contained low, but
detectable, levels of Monacolin K.
[588]
I will begin with the results from Dr. Harding’s tests on two commercial
samples of Red Yeast Rice. Even if those samples did contain measurable
quantities of lovastatin (Monacolin K) (which I am prepared to accept), these samples
were produced well after the priority date of the invention of the '380 Patent.
They tell us nothing about the existence of lovastatin in Red Yeast Rice at any
time prior to 1979. Moreover, beyond representations on the packaging, we have
no evidence as to how these commercial products were produced and whether the lovastatin
found in the samples resulted from traditional methods of fermentation.
[589]
As described by Dr. Harding, the modern techniques of producing Red Yeast
Rice would differ from traditional methods:
The modern techniques
would employ things like flasks and control environments. Traditionally these
things were fermented in boxes or bamboo plates, containers, in open areas, and
the temperature was controlled either by fanning or ensuring no direct
sunlight, these sorts of things, but also by mixing the rice to uniformly
distribute the yeast to get the full colour and the final product through every
kernel of rice, but it's also to reduce the temperature.
[590]
In his Expert Report, Dr. Harding also notes that, while traditional and
modern production practices are quite similar, “modifications [have] been
introduced into modern production to enhance the production of the desired
metabolites” (Harding Expert Report, Exhibit 115, para. 28).
[591]
Dr. Havel, an expert presented to the Court by Merck, also commented that:
Samples collected in
January to April 1998, 12 years after the Negishi paper, may have been produced
under conditions which were not, I believe, we can consider truly traditional
red yeast rice on or before 1980.
[592]
In my view, Dr. Harding’s opinion that the two samples he tested were
prepared in accordance with traditional methods is speculative. The results of
Dr. Harding’s tests on two Red Yeast Rice samples cannot be used to reliably
demonstrate that, before the priority date, Red Yeast Rice contained lovastatin.
[593]
I have similar difficulties with the literature referenced by Dr. Harding
in his Expert Report. One article was cited as Negishi et al, “Productivity of
Monacolin K in the genus Monascus species” (referred to as Negishi). Negishi
reports having tested 124 Monascus strains in total. Almost all of the
species were isolated from several different red yeast food products. Negishi
used modern techniques to carry out the experiments, and found that 17 strains
of five species were capable of producing Monacolin K. The initial problem that
I have with this paper is that it post-dates, by many years, the priority date
of the '380 Patent. I have no information on how the Red Yeast Rice samples
were collected and whether these same products would have been available in the
years before 1979.
[594]
Another reference by Dr. Harding was a 2001 journal article by Heber et
al, entitled “An Analysis of Nine Proprietary Chinese Red Yeast Rice Dietary
Supplements: Implications of Variability in Chemical Profile and Contents” (referred
to as Heber). Heber purchased nine different commercially available dietary Red
Yeast Rice supplements and found total Monacolin K content from 0% to 0.58%.
While the paper may demonstrate that a few samples of Red Yeast Rice products
collected and tested for purposes of the paper contained measurable amounts of Monacolin
K, they do not establish that lovastatin existed in Red Yeast Rice prior to
1979.
[595]
The only possible link between testing that occurred after 1979 and the
potential for the production of lovastatin from Red Yeast Rice prior to that
date would be the production methods. Dr. Harding appears to base his overall
opinion that traditional Red Yeast Rice has contained Monacolin K throughout
the centuries on the observed similarities between traditional and modern
production methods. If it can be established that the production methods used
in both periods were the same, it may be possible to extrapolate post-1979
results to pre-1979 samples of Red Yeast Rice.
[596]
Dr. Harding testified in cross-examination that, of the seven pieces of
prior art cited in his report, only two dealt with the fermentation of Red Yeast
Rice prior to 1979.
[597]
However, upon closer examination, there were significant differences in
the traditional methods of producing Red Yeast Rice and those that would
produce lovastatin, including:
·
traditional Red Yeast Rice was most likely be fermented at temperatures in
excess of 30ºC where lovastatin cannot be produced;
·
there is an inverse relationship between the amount of pigment-producing
ability and the amount of lovastatin-producing ability in Red Yeast Rice; and
·
the modern strains of Red Yeast Rice provide no information about whether
the traditional Red Yeast Rice produced lovastatin.
The evidence is strong that the
production of lovastatin is very dependent on the temperature of fermentation.
Although Negishi reported that 17 of 50 samples of the Monascus species
produced Monacolin K, they also reported that Monacolin K was produced only
when the fungi were grown at 25ºC. As observed by Negishi, “under conditions of
incubation at 30 to 37ºC, those Monacolin K-producing strains lost their
capability of producing Monacolin K . . .”. The '380
Patent teaches the reader to incubate the A. terreus fungus at 28°C, a temperature at which Monascus would likely be
incapable of producing lovastatin.
[598]
In sum, I am not persuaded that the evidence demonstrates that lovastatin
was contained in Red Yeast Rice prior to the priority date. It follows that
Apotex has not satisfied its burden of demonstrating that Red Yeast Rice
anticipates lovastatin.
(e) Disclosure of lovastatin in Red Yeast Rice
[599]
In the event that I am wrong in this conclusion, I will assume that
lovastatin was contained in at least some samples of Red Yeast Rice prior to
the relevant date. The question is whether this satisfies the requirement of
disclosure. In my view, as discussed in the following, Apotex has not persuaded
me that the subject matter of the invention - lovastatin – was disclosed to the
public by Red Yeast Rice.
[600]
The evidence demonstrates that, if Red Yeast Rice contained lovastatin, it
did not do so under all conditions of fermentation. That is, prior to June 15,
1979, not every Red Yeast Rice product contained lovastatin. Further, it is
also evident that persons who produced or used Red Yeast Rice were unaware that
it contained lovastatin.
[601]
Apotex argues that Calgon Carbon Corp. v. North Bay (City), 2008
FCA 81, 64 C.P.R. (4th) 337 at paragraph 8 and Baker Petrolite
Corp. v. Canwell Enviro-Industries Ltd., 2002 FCA 158, 17 C.P.R. (4th)
478 at paragraphs 35, 42 [Baker Petolite] stand for the proposition that
the extent or duration of the prior use or sale is not important; that
disclosure to “even one member of the public” destroys the novelty of a
chemical product. Thus, Apotex submits, the existence of lovastatin, in even
one sample of Red Yeast Rice, as of the priority date would be anticipatory of
the compound claimed in the '380 Patent.
[602]
I disagree with Apotex’s assertion that the existence of lovastatin in
even one sample of Red Yeast Rice is sufficient to demonstrate disclosure. I
think that Apotex confuses “disclosure to one member of the public” as stated
in Baker Petrolite, above, at paragraph 42, with “one disclosure”. If
the prior art invariably or predictably discloses the compound, there may well
be anticipation. However, where the existence of the compound alleged to be
anticipatory cannot be reasonably or consistently predicted from a large
universe of possibilities, I cannot see how this could possibly meet the test
for disclosure. Anticipation must be more than an accidental presence of a
compound.
[603]
Not only does Apotex’s argument appear illogical to me, it is not
supported by the jurisprudence. In Abbott Clarithromycin (FCA),
above, at paragraph 22, the Court of Appeal considered the argument of Abbott
that a skilled person must have certain knowledge regarding the prior art.
Abbott argues that a
person skilled in the art who heated clarithromycin Form I by the known
technique would not and could not know that clarithromycin Form II had been
created, unless they also knew that the heating process had to be stopped
before the substance reached its melting point at 225ºC. In my view, the
absence of that knowledge is legally irrelevant. The undisputed evidence is
that clarithromycin Form II would have been present if the heating
technique had been followed. There were well established analytical techniques
that would have disclosed its presence if anyone had cared to look at the
appropriate moment. [Emphasis added.]
[604]
In the case before me, I have no such evidence that lovastatin would
have been present in Red Yeast Rice. Indeed the evidence, as disclosed in
the literature cited (for example, Negishi and Heber), is that lovastatin was
present only in a very few samples.
[605]
Apotex also argues that the fact that no one, including the skilled
person, ever recognized that Red Yeast Rice contained lovastatin is not
relevant. Apotex submits that the Federal Court of Appeal, in Abbott
Laboratories v. Canada (Minister of Health), 2006 FCA 187, 56 C.P.R. (4th)
387 at paragraphs 15-23, refers to the principle, outlined by Lord Hoffmann in BV v. Smithkline Beecham plc, [2005] UKHL 59,
[2006] 1 All ER 685 [Synthon BV], that a patent is disclosed even though the author or maker of the prior art was not aware that he was
doing so. Apotex refers to Synthon BV, above, at paragraphs 22-23
for support on this point.
[606]
I agree with Apotex that the cited passage of the House of Lords decision
in Synthon BV, above, does contain a reference to the fact that an
awareness of infringement is not a requirement. However, an examination of the
entire passage clarifies the context in which those remarks were made. As
stated by Lord Hoffmann in Synthon BV, above, at paragraphs 22-23:
. . . the matter
relied upon as prior art must disclose subject matter which, if performed,
would necessarily result in an infringement of the patent. . . . But patent
infringement does not require that one should be aware that one is infringing:
“whether or not a person is working [an] ... invention is an objective fact
independent of what he knows or thinks about what he is doing”: Merrell Dow
Pharmaceuticals Inc v H N Norton & Co Ltd [1996] RPC 76, 90. It follows
that, whether or not it would be apparent to anyone at the time, whenever
subject-matter described in the prior disclosure is capable of being performed
and is such that, if performed, it must result in the patent being infringed,
the disclosure condition is satisfied. The flag has been planted, even though
the author or maker of the prior art was not aware that he was doing so.
Thus, in Merrell
Dow, the ingestion of terfenadine by hay-fever sufferers, which was the
subject of prior disclosure, necessarily entailed the making of the patented
acid metabolite in their livers. It was therefore an anticipation of the acid
metabolite, even though no one was aware that it was being made or even that it
existed. But the infringement must be not merely a possible or even likely
consequence of performing the invention disclosed by the prior disclosure. It
must be necessarily entailed. If there is more than one possible consequence,
one cannot say that performing the disclosed invention will infringe. The flag
has not been planted on the patented invention, although a person performing
the invention disclosed by the prior art may carry it there by accident or (if
he is aware of the patented invention) by design. Indeed, it may be obvious to
do so. But the prior disclosure must be construed as it would have been
understood by the skilled person at the date of the disclosure and not in the
light of the subsequent patent. As the Technical Board of Appeal said in T/396/89
UNION CARBIDE/high tear strength polymers [1992] EPOR 312 at para 4.4:
“It may be easy, given
a knowledge of a later invention, to select from the general teachings of a prior
art document certain conditions, and apply them to an example in that document,
so as to produce an end result having all the features of the later claim.
However, success in so doing does not prove that the result was inevitable.
All that it demonstrates is that, given knowledge of the later invention, the
earlier teaching is capable of being adapted to give the same result. Such an
adaptation cannot be used to attack the novelty of a later patent.”
[607]
An understanding of this passage demonstrates that Apotex has selectively
read the comments of Lord Hoffmann, omitting a key consideration. I agree with
Apotex that an awareness that the prior art discloses the anticipatory
invention is not a requirement of disclosure. Justice Hughes made that point in
Abbott Clarithromycin (FC), above, at paragraph 75. However,
while supporting the argument of Apotex, the passage from Synthon BV, above,
makes it very clear that the disclosure requirement is only met where
infringement must occur when the prior art is practised. Paraphrasing
the words of Lord Hoffmann, the flag of anticipatory disclosure is not planted
on the patent in question unless the prior disclosure would necessarily result
in an infringement of the patent.
[608]
Applied to the situation before me, Red Yeast Rice only satisfies the
disclosure requirement of the test for anticipation if it can be said to
necessarily produce lovastatin. As we know from the evidence, that is
definitely not the case. Indeed, the evidence before me is that the existence
of lovastatin in Red Yeast Rice would be a very rare event.
[609]
In conclusion on this issue, I do not accept Apotex’s submission that
lovastatin was anticipated by Red Yeast Rice.
XI. Conclusion
[610]
As noted at the beginning of these reasons, this litigation was subject to
the Bifurcation Order. Thus, the matter proceeded to trial without requiring
the parties to adduce evidence at trial on any issue of fact pertaining to the
following:
1.
the extent of infringement, if any, by the Defendants of the ‘380 Patent;
2.
the amount of damages suffered, if any, by the Plaintiffs as a result of
any such infringement; or
3.
the amount of profits earned by the Defendants from any such infringement.
[611]
According to the terms of the Bifurcation Order, the determination of
whether the Plaintiffs are entitled to elect to recover profits was to be
determined by the trial judge. In final arguments, Merck submitted that it
wished to elect to recover profits. Apotex objected.
A. Damages or Profits
[612]
Once a patentee has successfully demonstrated infringement, the Court has
the discretion to grant the patentee's choice of remedies – either damages (pursuant
to s. 55 of the Patent Act) or an accounting of profits (pursuant to s.
57 of the Patent Act). Merck wishes to elect an accounting of profits
and asks this Court to so direct. Apotex argues that I should not exercise my
discretion in this case.
[613]
While both damages and accounting of profits are intended to provide
compensation to a wronged plaintiff, the fundamental principles underlying the
two remedies and the practical considerations are substantially different.
[614]
The object of an award of damages is to make good any loss suffered by the
plaintiff as a result of the defendant's infringement of the patent.
Quantification of the award is based on the losses suffered by the plaintiff;
any gains realized by the defendant because of its wrongdoing are not relevant.
On the other hand, an accounting of profits is based on the premise that the
defendant, by reason of its wrongful conduct, has improperly received profits
which belong to the plaintiff. The objective of the award is to restore those
actual profits to their rightful owner, the plaintiff, thereby eliminating
whatever unjust enrichment has been procured by the defendant. Calculation is
based on the profits wrongfully gained by the defendant; any other losses
suffered by the plaintiff are irrelevant.
[615]
An accounting of profits is not an easy calculation. As was stated by the
late Justice Paul Rouleau, of this Court, when speaking about such an
accounting in Beloit Canada Ltd. v. Valmet-Oy. (1993), 55 C.P.R. (3d)
433 at para. 3, [1994] F.C.J. No. 682 (F.C.T.D.)(QL), rev'd in part 61 C.P.R.
(3d) 271, [1995] F.C.J. No. 733 (QL)(F.C.A.), leave to appeal to S.C.C.
refused, [1995] S.C.C.A. No. 388 (QL), 64 C.P.R. (3d) vi:
This was undoubtedly a
most expensive, lengthy and difficult reference and one which clearly
underlines the pitfalls of granting the remedy of an accounting of profits
other than in exceptional and appropriate circumstances and after due
deliberation by the court.
[616]
In spite of practical difficulties, the Federal Court of Appeal in Beloit
Canada Ltd. v. Valmet Oy (1992), 45 C.P.R. (3d) 116 at para. 10, [1992]
F.C.J. No. 825 (QL), stated that it could:
...see no reason in
principle why a patentee, whose property has been wrongly appropriated through
infringement, should not recover all the profits, direct and indirect, derived
by the infringer from his wrongful infringement.
[617]
It is necessary for a party seeking an equitable remedy, such as profits,
to show some basis for the exercise of equity (Janssen-Ortho Inc. v.
Novopharm Ltd., 2006 FC 1234, 57 C.P.R. (4th) 58 at para. 132,
aff'd 2007 FCA 217, 59 C.P.R. (4th) 116 leave to appeal to S.C.C.
refused, [2007] S.C.C.A. No. 442 (QL), 383 N.R. 397 (note); Servier FC,
above, at para. 507).
[618]
Merck submits that it can demonstrate a basis for an election of profits.
Specifically, it puts forward the following factors:
·
The Plaintiffs have not committed any inequitable conduct which would
disentitle them to equitable relief.
·
The Plaintiffs did not delay in commencing the litigation. The
infringement action was commenced on June 12, 1997, within 3 months of Apotex
Inc. receiving an NOC for Apo-lovastatin.
·
The Defendants could not have doubted that the Plaintiffs would pursue an
infringement action given the lengthy history of prior lovastatin litigation.
·
The reasons why the litigation took years to prosecute are not, Merck
submits, reasons to disentitle the Plaintiffs to equitable relief:
○
the long and complicated facts of the cases: we are dealing with a process
patent, where the acts of infringement are conducted in secret, requiring the
Plaintiffs to attempt to prove infringement through a long discovery process
and repeated motions for productions;
○
AFI’s history of use of Aspergillus terreus to make lovastatin
going back to 1991 to 1999;
○
the transfer of Aspergillus terreus and AFI‑1 technology to
Blue Treasure in 1995 and its use;
○
the filing of a statement of claim in T-1169-01;
○
Apotex Inc.’s refusal to agree to consolidate the two proceedings for
discovery meaning that the Plaintiffs had to repeat to a large extent the
discovery from the infringement action;
○
the patent expiry on January 31, 2001, thereby capping the period of
infringement as of that date; and
○
two years of the delay due to the length of the trial requested.
·
Delay in the action is a factor that is more related to the complicated
facts of the proceeding rather than the diligence of the Plaintiffs.
[619]
While I agree with Merck that it has not committed any inequitable conduct
that would disentitle them from the equitable remedy of profits, other factors
weigh against such a remedy in this case.
[620]
A factor that causes me serious concern is the time that this matter took
to come to trial. Merck attempts to distance itself from any decisions that
resulted in the delay of this trial for almost thirteen years. I cannot accept
that the Defendants and the Federal Court bear all of the responsibility for
the delay. The consequence of this delay is, inevitably, that reaching back to
the period between 1997 and 2001 to assess Apotex profits would be exceedingly
difficult.
[621]
The difficulty in assessing profits is further exacerbated by the
complexities of the commercial arrangements that involved not only AFI and
Apotex Inc., but also Blue Treasure and Biogal. Merck consented to the
settlement involving Biogal’s interests on May 28, 2010.
[622]
An additional layer of complexity comes from the fact that Merck does not
assert that all of the lovastatin that was produced during the 1997 to 2001
period infringed the '380 Patent. On the
Canadian‑produced material by AFI, except for the batch referred to as CR0157,
there is no assertion of infringement. I have also concluded that Merck has not
made out its case for infringement with respect to all of the Blue Treasure
productions.
[623]
Dissecting Apotex’s profits to account for the Biogal settlement and the
non-infringing production would be a complex undertaking.
[624]
Balancing the factors in this case, I am not persuaded that I ought to
exercise my discretion and permit the Plaintiffs to elect an accounting of
profits. The Plaintiffs will be entitled to their damages. Specifically, a
hearing under ss. 107 and/or 153 of the Federal Courts Rules, SOR/98-106
[Federal Courts Rules] shall be conducted to determine: the extent of
infringement by the Defendants of the '380 Patent;
and, the amount of damages suffered by the Plaintiffs as a result of such
infringement.
B. Exemptions from Liability
[625]
Related to the issue of damages is the question of whether any volumes of
lovastatin produced by Apotex should be exempt from a finding of infringement.
Apotex relies on s. 55.2(1) of the Patent Act (post October 1, 1989) to
submit that it should not be held liable for any infringement relating to its
experimental and regulatory uses of lovastatin.
[626]
Exemptions from liability are founded in s. 55.2(1) of the Patent Act
(post October 1, 1989). That provision states that:
|
55.2(1) It is not an infringement of a patent for any
person to make, construct, use or sell the patented invention solely for uses
reasonably related to the development and submission of information required
under any law of Canada, a province or a country other than Canada that
regulates the manufacture, construction, use or sale of any product
|
55.2(1) Il n'y a pas contrefaçon de brevet lorsque
l'utilisation, la fabrication, la construction ou la vente d'une invention
brevetée se justifie dans la seule mesure nécessaire à la préparation et à la
production du dossier d'information qu'oblige à fournir une loi fédérale,
provinciale ou étrangère réglementant la fabrication, la construction, l'utilisation
ou la vente d'un produit.
|
[627]
Apotex may claim an exemption from liability for certain amounts of the
infringing product, provided that it can satisfy its burden to demonstrate that
the product was used for permitted purposes (such as obtaining regulatory
approval or to comply with regulations) (Merck & Co. v. Apotex Inc.,
2006 FC 524, 53 C.P.R. (4th) 1, rev'd on other grounds 2006 FCA 323,
55 C.P.R. (4th) 1 leave to appeal to S.C.C. refused, [2006] S.C.C.A.
No. 507 (QL), 370 N.R. 400 (note)).
[628]
The Canadian Food and Drug Regulations, C.R.C. 1978, c. 869 [Food
and Drug Regulations], and the United States Food Drug and Cosmetic Act,
21 U.S.C., as amended [Food Drug and Cosmetic Act], required Apotex Inc.
to retain and test samples of lovastatin on a routine basis. Apotex submits
that its testing and retention of lovastatin falls within the scope of
subsection 55.2(1) of the Patent Act and the common law exception and is
therefore exempt from infringement.
[629]
Apotex submits that the evidence clearly shows that, as required by the
Canadian Food and Drug Regulations, above, and the regulations under the
U.S. Food Drug and Cosmetic Act, above, Apotex: (i) acquired and used
bulk lovastatin, and formulations incorporating bulk lovastatin, for the
purpose of obtaining permission to sell lovastatin containing pharmaceutical
products in Canada and the United States; (ii) carried out in process quality
control sampling of lovastatin formulations; and (iii) retained samples of bulk
and finished dosage forms of lovastatin.
[630]
Mr. Barber, the Manager of the Formulations Department at Apotex Inc., and
Ms. Copsey, Manager of Packaging and Director of Commercial Lab Operations at
Apotex Inc., explained in detail how Apotex Inc. used lovastatin for these
purposes. The business records of Apotex Inc. adduced at trial reflect those
uses. Apotex submits that its use of lovastatin for these experimental and
regulatory purposes is exempt from infringement under s. 55.2(1) and (6) of the
Patent Act (post October-1989) and the common law.
[631]
Mr. Fahner was the Vice-President of Finance at Apotex Inc. during the relevant
times. He compiled charts identifying the quantities of lovastatin from each
lot of bulk lovastatin and finished dosage batch that were used by Apotex Inc. for
each of the purposes described by Mr. Barber and by Ms Copsey, namely, the research
and development work in preparation of the submission batches, the retention of
API samples and the process sampling and finished goods retention. Apotex Inc.
submits that the evidence establishes that the following quantities of
lovastatin were used by Apotex Inc. for
regulatory and experimental purposes and should be exempt from any finding of
infringement:
|
Use of
lovastatin by Apotex Inc.
|
Total Quantity (in kg)
|
|
Research and Development
|
59.1111
|
|
Reserve Samples (API)
|
22.0986
|
|
In Process Samples
|
6.58078
|
|
Finished Goods Retained Samples
|
4.2654
|
[632]
I do not understand Merck to be objecting to the exemption of these Apotex
Inc. volumes from a finding of infringement. I am satisfied that the volumes
set out in the above table are exempt from any finding of infringement.
[633]
AFI also conducted research and development work involving the use of the
micro-organism Aspergillus terreus after the assets of ABI were acquired
in July 1991. Initially, the work was a continuation of the work commenced by
ABI in 1988 and was specifically related to obtaining a compulsory licence.
Subsequently, the work involved the research and development of Aspergillus
terreus for regulatory submissions and for eventual commercialization of Aspergillus
terreus lovastatin. In the course of these activities, ABI, and
subsequently AFI, manufactured 6.9 kg of Aspergillus terreus lovastatin.
This was supplied to Apotex Inc. and used for research and development purposes
and for regulatory submissions. AFI submits that this work is exempt from
infringement. I agree.
[634]
In 1998 and 1999, AFI ran a few fermentations using Aspergillus terreus
at the 14,000 litre scale for research and development purposes directed to
market readiness upon expiry of the '380 Patent. A total of 13.45 kg was
manufactured from these runs. The Defendants assert that this work is exempt
from infringement. In June 2002, this material was supplied to Brantford
Chemical Inc., a sister company to Apotex Inc., for development purposes in an
effort to make simvastatin, another statin that is not the subject of this
litigation. Apotex Inc. submits that there was no evidence at trial that the
13.45 kg was ever used to manufacture tablets for sale in Canada, or elsewhere.
This work is exempt from infringement. I agree.
[635]
Merck objects to an exemption from liability for any amounts of AFI-1
lovastatin made in Winnipeg during the period 1993 to 1999. In their view, these
volumes clearly infringed the '380 Patent. Moreover, through the transfer of the AFI-1
technology to Blue Treasure for commercial purposes, the infringement resulted
in a loss of Merck’s right to the “full enjoyment of the monopoly” (Monsanto,
above, at para. 34). Accordingly, Merck argues, these volumes should not be the
subject of any “fair dealing” exemption. In addition, Merck submits that the
liability should extend to all 296.6 kg of AFI-1 lovastatin that were allegedly
made by Blue Treasure.
[636]
I am not prepared to make this link. In light of the Supreme Court of
Canada’s decision in Micro Chemicals Limited v. Smith Kline & French
Inter-American Corp., [1972] S.C.R. 506, 2 C.P.R. (2d) 193 [Micro
Chemicals], Merck’s allegation is without foundation. In the case at bar,
AFI was doing exactly what the defendant did in Micro Chemicals. AFI was
carrying out research for the purposes of improving its Aspergillus terreus
process to ensure that the process could be used on a commercial scale. This is
the type of activity that the Supreme Court held was exempt from infringement.
The supply of the developed technology to Blue Treasure, who was permitted to
utilize the AFI-1 process to make lovastatin for sale outside Canada, was not,
in the circumstances, an act of infringement.
[637]
In sum, I am satisfied that neither the 6.9 kg nor the 13.45 kg of Aspergillus
terreus lovastatin referred to above should attract liability. Moreover, I
am not prepared to conclude that the alleged 296.6 kg of AFI-1 lovastatin that
was made using the transferred AFI-1 technology infringed the '380 Patent.
C. Conclusion
[638]
In conclusion, the action of the Plaintiffs will succeed, in part, and
they will be entitled to an order for the recovery of damages sustained as a
consequence of the Defendants’ infringement of the '380 Patent by the
following:
1.
all Apo-lovastatin product that was produced by AFI from AFI batch CR0157;
and
2.
the 294 batches of lovastatin produced by Blue Treasure in China after
March 1998 and imported into Canada.
[639]
Damages and the extent of infringement will be determined by way of a
reference pursuant to Rule 153 of the Federal Courts Rules and in
accordance with the Bifurcation Order dated November 14, 2003. The parties will
be permitted to address the specific terms of a reference as to damages, by way
of written submissions to be served and filed within 60 days of the date of the
Judgment. The parties will have a further 15 days to serve and file reply
submissions.
[640]
Merck will also be entitled to an order for prejudgment interest pursuant
to ss. 36 and 37 of the Federal Courts Act, S.C. 2002,
c. 8.
[641]
The counterclaims of the Defendants will be dismissed. Specifically, I
conclude that the '380 Patent was valid and that Merck & Co. has standing
to bring this action.
[642]
The question of costs was not addressed by the parties in their final
submissions. Obviously, Merck, as the successful party will be entitled to
costs, although the amount of those costs should reflect the specific
circumstances of this trial. The parties will be given a period of time to
attempt to resolve the issue of costs among themselves. I sincerely hope that
the direction of this Court is not required. However, in the event that the
parties cannot agree on costs, they may serve and file submissions, not to
exceed ten pages in length, within 60 days of the date of the Judgment. The
parties will have a further 15 days to serve a reply, any such reply not to
exceed five pages.
POSTSCRIPT
[1]
These
Reasons for Judgment are un-redacted from confidential Reasons for Judgment
which were issued on December 9, 2010 pursuant to the Direction dated December
9, 2010.
[2]
The Court
canvassed counsel for the parties whether they had concerns if the reasons were
issued to the public without redactions. On December 15, 2010 and December
17, 2010, in separate letters, the parties advised that there are no portions
of the confidential Reasons for Judgment that should be redacted.
“Judith A.
Snider”
Ottawa, Ontario
Public Reasons – December 22, 2010
Confidential Reasons - December 9, 2010
Email this Content