The Regeneron patent EP1360287 relates to methods of producing transgenic mice possessing human antibody genes. The divisional patent EP2264163 claims the mouse itself. Historically, the development of therapeutic antibodies relied on the cloning of specific human antibody genes into a mouse. The Regeneron transgenic mice (VelocImmune) produce antibodies having a human variable region and a mouse constant region, which respond naturally to antigenic challenge and thus mirror the normal immune reaction. This facilitates production of a diverse spectrum of antibodies against the same target within a single mouse.
The Regeneron patents particularly relate to a method for genetically modifying the antibody variable regions of a mouse cell by replacement of the mouse antibody variable genes with the equivalent human genes. The resulting mouse cells contain a human antibody light chain (VJ) or heavy chain (VDJ) variable region that undergoes natural homologous recombination during B-cell development. The method involves cloning a large (>20 kb) genomic fragment, containing one or more human V, J, (and optionally D) regions, into a targeting vector (LTVEC), introducing this vector into a mouse cell and screening for successful insertion of the human V, D and/or J, using a modification of allele (MOA) assay. The specification includes an example (Example 3) describing in situ replacement of the mouse variable region (VDJ/VJ) genes with their human counterparts.
The VelocImmune mouse has proved exceedingly commercially valuable to Regeneron over the past few decades. Regeneron has brought a number of therapeutic antibodies to market using the platform and have many others currently undergoing clinical trial. In 2007, Regeneron agreed a six-year US$120 million non-exclusive licence with AstraZenca for use of the technology, under which Regeneron receives royalties on the sale of any products produced using the platform. Astellas Pharma, one of the largest pharmaceutical companies in Japan, has also agreed a non-exclusive license for use of the VelocImmune technology until 2023.
UK High Court
In the UK, Regeneron brought infringement proceedings against Kymab and Novo, for infringement of the EP(UK) patents. The defendants counterclaimed that the patents were invalid on the grounds that the claimed invention was not sufficiently disclosed and lacked novelty and inventive step. Mr Justice Carr rejected the novelty and inventive step objections, but agreed with the defendants regarding the lack of sufficiency, on the basis of which he revoked the patents.
Mr Justice Carr found that the claims were insufficient in view of expert evidence that the method provided in Example 3 of the specification would not have worked in the hands of the skilled person at the priority date (Regeneron’s own work after the priority date confirmed this). The method claim specifies “in situ replacement of V, D and J gene segments of the endogenous locus with orthologous human V, D and J gene segments”. Mr Justice Carr interpreted this as covering the deletion of mouse sequences of at least one V segment and all the D and J segments (>100 kb) and the insertion of the equivalent human regions (>75kb). On the basis of expert evidence that the insertion and deletion of such large pieces of DNA was not possible at the time of the invention, Mr Justice Carr thus concluded that the claimed method and mouse were insufficiently disclosed.
The Court of Appeal – Insufficiency
In their appeal, Regeneron did not challenge the finding that a skilled person would not have been able to delete 100kb and insert 75kb of DNA. Instead they argued that a skilled team would have made adjustments to the method of Example 3, so as to reduce the size of the inserts to a manageable length, by reducing the number of V, D and J regions and removing unnecessary intergenic regions of sequence, the “minigene” approach. The deletion of the mouse regions could be carried out in a single step, and additional human V regions (10kb) added as necessary.
Kymab argued that Regeneron had not previously raised the minigene argument and that it was now too late for it be submitted. The Court of Appeal did allow submission of Regeneron’s arguments on the basis that Regneron had, in fact, previously contended that the claims did not necessitate insertion of the entire human locus or deletion of the entire mouse locus, and that shorter sequences in the form of minigenes could be used. Despite this, Mr Justice Carr did not deal with the issue of minigenes in his judgement. It was of “considerable concern” to the Court of Appeal that Regeneron did not raise this when the draft judgement was issued. The Court of Appeal nevertheless allowed Regeneron’s submissions, given that the evidence was not extensive and was primarily given by Kymab’s experts.
In contrast to Mr Justice Carr, the Court of Appeal found that the concept of the minigene was common general knowledge at the priority date, and that a skilled team would have appreciated that the claimed method could be achieved by producing a VDJ minigene without undue effort. This conclusion was based on the previous testimony of Kymab’s expert witnesses Professor Stewart and Professor Howard.
In the context of a discussion of the prior art, Professor Stewart stated that a skilled person would have appreciated at the priority date that the largest replacement that could be made by homologous recombination would be less than 20 kb and that replacement of the mouse locus with the human locus would therefore have required more than 100 rounds of homologous recombination and taken years to complete, representing undue effort. However, Professor Stewart also stated that “the only way it would have been feasible to create [the mouse] would be to use a small minigene construct up to approximately 20 kb in length containing a subset of the…gene segments”. Professor Howard similarly stated that the skilled person could have engineered a minigene construct including rearranged V, D and J segments.
The Court of Appeal did not disagree with the trial judge that Example 3 would not have worked, or that at the priority date a skilled person would not have known how to delete 100 kb of mouse sequence and insert 200-300kb of human sequence. However, the Court of Appeal noted that the skilled person is not bound to carry out the invention precisely as described, and can use common general knowledge to make obvious changes. Furthermore, the expert testimony indicated that a skilled person would have recognized Example 3 to be considerably challenging, and would have considered it obvious to shorten the inserts. Finally, the court of appeal pointed out that “a patent does not cease to be sufficient simply because the specification promises too much”.
In July 2017, US Court of Appeals for the Federal Circuit upheld the decision that the US patent was unenforceable (full judgement here). The decision affirmed the patent could not be enforced because of “Regeneron’s inequitable conduct” during prosecution, by intentionally withholding references from the USPTO.
The Court of Appeal brings the UK courts in line with the EPO for this case. It is worth noting however, that the UK courts had the benefit of more expert evidence than the EPO. Interestingly, the evidence on which the Court of Appeal based their decision was that provided by Kymab’s experts for the purpose of novelty. A reminder of the trap to be avoided between demonstrating that a prior art document is enabling and thus destroys the novelty of a claim, and subsequently attempting to prove that the same claim is insufficient.
This article is for general information only. Its content is not a statement of the law on any subject and does not constitute advice. Please contact Reddie & Grose LLP for advice before taking any action in reliance on it.