Last year, the Australian Federal Court issued its decision concerning the Prevnar 13 vaccine in Merck Sharp & Dohme Corporation v Wyeth LLC  FCA 1477 (Prevnar 13 AU). This year, the Federal Court of Canada decided very similar questions in Merck Sharp & Dohme v Wyeth Canada Inc 2021 FC 317 (Prevnar 13 CA). While the cases are interesting for a range of reasons, a comparison between the two decisions shows different approaches to the construction of “comprising” when used in claims which warrant consideration.
Streptococcus pneumoniae causes serious bacterial infections and addressing it has been a challenge for many years and, for the companies involved in this litigation, since the 1980s. Immunisation can address this issue. Wyeth had immunised over 1.5 billion people around the world within a decade covered by the relevant patents which represented a significant improvement on earlier vaccines, such as Prevnar 7 and Prevnar 9, which had lesser serotype coverage, i.e. less effective against pneumococcal disease.
Thus, Wyeth’s patents concerning Prevnar 13 were of great value and contentious, leading to litigation with Merck around the world as Merck wished to commercialise its own 15-valent pneumococcal polysaccharide-protein conjugate vaccine.
The key issue was a relatively simple one of construction – the gateway to analysis in any patent litigation.
Questions of Construction
Claim 1, from the Australian and Canadian composition patents, read (in part):
A multivalent immunogenic composition, comprising 13 distinct polysaccharide-protein conjugates, together with a physiologically acceptable vehicle, wherein each of the conjugates comprises a capsular polysaccharide from a different serotype of Streptococcus pneumoniae.
So arose an old question in patent law, the bounds of the term “comprising”. Did this claim intend:
- that the claim be limited to a composition consisting of 13 distinct polysaccharide-protein conjugates, or
- that the claim encompassed a composition including 13 or more distinct polysaccharide-protein conjugates?
Unsurprisingly, Merck argued for the first construction (its proposed vaccine had 15 serotypes). The invention was at most a 13-serotype vaccine and, if it was incorrect on that, the claim was overbroad and not supported (from the perspective of Australian patent law).
Wyeth, on the other hand, argued that its patents provided a platform to make polysaccharide conjugate vaccines comprising 13 or more serotypes of S. pneumoniae. The Prevnar 13 vaccine was the result of years of experimental work and it was the ‘gold standard’ of the industry. The skilled person would appreciate that, through the expression “comprising 13”, the composition patent was aimed at providing such a platform broadening the range of serotypes and coverage of pneumococcal infections in a “next generation” of its development program. The skilled person would have used the platform to add other serotypes in a future vaccine. Wyeth argued that the term “comprising” should be construed in an inclusive sense, meaning “including but not limited to”. This represents a position where the claims, and by extension the specification, simply provide a guide to – rather than a boundary fence around – the protected invention.
Merck argued that the word “comprising” allows for more elements in the composition, such as acceptable physiological vehicles and/or adjuvants. However, the claimed compositions were only to polysaccharide-protein conjugates of each of the chosen 13 serotypes. This construction, which gave a common sense and practical result, was the only one consistent with the description in the specification.
The Australian Decision
In Australia, the Court accepted Wyeth’s construction. The words “comprising” and “comprises” were clearly defined in the specification in an inclusive sense. The only exception was if the context required otherwise.
The scope of claim 1 was to be defined by what the composition includes not what it consists of so, as a matter of impression, any composition containing 13 serotypes (e.g. 15 serotypes) would be within the scope of claim 1 unless the language of the claims supported an alternative construction. The specification was not sufficient to resolve this question.
The Court also dismissed Merck’s argument that, if Wyeth had wanted to claim immunogenic compositions with more than the 13 chosen serotypes, it could have done so by using the words “at least” before listing the serotypes but did not. However, the Court had to determine what the claim wording was not what it might have been. Nevertheless, there is a message here for the patent drafter.
On interpretation of the specification, Burley J appeared to be moved by a reference that the invention provides generally a multivalent composition comprising 13 serotypes. No maximum number of potential serotypes was given by the specification.
Wyeth, therefore, succeeded in establishing that “comprising” was used in an inclusive sense, effectively meaning that claim 1 addressed compositions including at least 13 serotypes. Following resolution of another construction argument, a finding of infringement followed.
The Canadian Decision
In Canada, the Court accepted Merck’s construction. Claim 1 was limited to 13 serotypes. Whilst the word “comprising” was open-ended, inclusion of other elements required some justification as provided by the patent description. On a reading of the patent, with which the present writer respectfully agrees, there was sufficient basis for adding adjuvants, buffers and other excipients. However, there was no basis to go beyond 13 serotypes. Relevant parts of the description included the Summary of the Invention and the examples. All 16 examples referred only to the specific serotypes recited in the claims and no others.
The Court also observed that, if it sided with Wyeth on the construction of claim 1, it would have to choose a cap on the number of serotypes covered by the invention or that could be added using Wyeth’s “so-called ‘platform’”. It would also mean that the ambit of Wyeth’s monopoly could grow over the life of the patent as new serotypes were discovered. This would be contrary to fundamental principles of claim construction.
The “comprising” construction issue was resolved differently in Australia and Canada and provides guidance for those preparing and enforcing patents, especially patents in the pharmaceutical and immunology fields. The Australian decision certainly favours a broad construction of a claimed invention and, on this occasion, provides a result in which the drafting patent attorney might well be happy. On another occasion, such a reading might cause a collision between the claim and prior art but not this time.
However, the Canadian decision better reflects what the outcomes for drafting should be where the briefing is of an invention having broader reach, i.e. where an increasing number of serotypes, and better coverage of the infectious disease (the platform theory) could be achieved. It would be a straightforward matter, as Merck argued, to recite “at least 13” and, while Burley J is, with respect, correct that the Court must construe the words as they stand, the result is more generous than expectation. Another interesting point is that, certainly under current applicable law requiring support, rather than fair basis, another serious issue might have arisen for Wyeth, though plausibility might have saved the day there.