UK: House of Lords Clarifies Fundamental Principles of Patent Law

This article originally appeared in the December 2004/ January 2005 edition of Managing Intellectual Property

It is extremely rare for any patent case to make it to England’s highest court. Now, however, within a week, two decisions have been issued which cover three of the main aspects of patent law – infringement, obviousness and sufficiency – as well as redefining one important aspect of the law of novelty.

The decision in SABAF v Meneghetti is useful in reaffirming the traditional view of how obviousness is to be assessed, but is not particularly controversial – overturning a Court of Appeal decision which many thought to have been wrongly decided. On the other hand, the decision in Amgen v TKT is much more controversial. Although the Catnic principles which underlie purposive construction were affirmed and held to be in line with principles applied in continental European jurisdictions, the House of Lords raised the possibility of jettisoning the rigid formulaic approach of the so-called ‘Improver’ or ‘Protocol’ questions when deciding infringement in suitable cases. It also unexpectedly held one of the patent’s claims to be invalid based on insufficiency and made other obiter comments which will refocus patentees’ attention on the need to claim no more broadly than their invention really merits. In addition, the way the English Courts approach product by process claims was completely changed, bringing it into conformity with our European neighbours.

On the face of it, it is rather surprising that what started life as a rather routine patent infringement case should end up in the House of Lords. The principal cause of the controversy was the reaction of the Court of Appeal to the approach taken on obviousness by Laddie J at first instance. He took a short cut and did not apply in the usual methodical way the four-part test from Windsurfing¹. Instead, he applied the rather ancient principle of "mere collocation" as made famous in the sausage machines case². He said:-

He then concluded:-

The Court of Appeal criticised this approach saying that:-

Accordingly, the Court of Appeal reversed Laddie J’s decision, albeit that having held the patent valid, they found it not to be infringed by reason of them also overturning Laddie J’s finding that Meneghetti was responsible for the importation of the relevant product³.

However, the House of Lords has now reaffirmed Laddie J’s original decision finding the patent to be invalid for obviousness. In so doing they have confirmed two aspects of the law on obviousness. First, that the UK Courts should apply the Windsurfing test as originally formulated by Oliver LJ (as he then was) in 1985. But they also found that in applying the old law of collocation, Laddie J had, in fact, applied Windsurfing correctly. Giving the leading Judgment, Lord Hoffman said:-

In reaching the decision, the House of Lords considered the impact of the EPO Guidelines, again confirming the desire of the UK Courts to ensure, so far as possible, consistency in the application of the European Patent Convention (EPC). There can be no doubt, therefore, that the House of Lords views the Windsurfing test as consistent with the EPC. It must also follow that applying the law of collocation in the way in which Laddie J did in SABAF is also consistent with the EPC. This seems entirely correct at least so far as it concerns the way in which the EPO applies the EPC. Everyone considers that in cases where known products are combined, the EPO is customarily concerned to consider the interaction or synergy between the products. This is, of course, the flipside of a case of a mere collocation.

Will the Judgment of the House of Lords then result in more UK patents being held invalid for obviousness? The answer is almost certainly not. The facts of SABAF were relatively unusual, and the application of standard EPO principles would normally be expected to result in patents such as that in SABAF being refused during prosecution – the patent in suit in SABAF was, in fact, a GB not an EP patent.

In contrast to SABAF, Amgen was from the outset a patent case which raised interesting legal and factual issues that were destined to push the boundaries of a number of areas of patent law.

The patent in question was born in the biotechnology field and concerned expression of the gene sequence for a naturally occurring protein hormone, erythropoietin (epo), which regulates production of red blood cells in the body. Epo is administered therapeutically to treat conditions such as anaemia and has become a multi-billion dollar blockbuster drug. The Amgen patent was said to be the first document to disclose the epo gene sequence, which had long proved illusory. This difficulty in identifying the DNA sequence was caused by the short supply of natural epo which is expressed in such vanishing small quantities in the body that not enough had been isolated to identify its sequence. However, the discovery of a gene sequence is not in itself a patentable invention, and the claims therefore covered the use of DNA sequences for producing epo and like proteins, and in a round about way, the proteins so produced.

The fact that this area of technology had moved on added an extra dimension to the case. TKT’s method of gene activation used the gene sequence information from the patent but used technology which, when the patent draftsman put pen to paper, would have been thought not to work. Interesting legal issues of patent law therefore arose as to whether a patent’s claims could be construed purposively to cover methods not contemplated at the time of the invention, and, if so, whether such a specification could enable the invention sufficiently.

The UK litigation had been running for many years culminating in a trial in the Patents Court before Mr Justice Neuberger in the late-1990s. At that stage both TKT and Roche were defendants. Neuberger J held that Amgen’s Patent was partially valid (claim 19 being held insufficient) and ruled there had been infringement by Roche and TKT. Soon after the first instance judgment Roche, which used a process more akin to Amgen’s, settled the litigation. TKT appealed the decision to the Court of Appeal. There, the decisions on invalidity and infringement were reversed, with the Court of Appeal ruling that Amgen’s Patent was valid but, whilst the claims had a wide scope of protection, they were not infringed by TKT.

The Court of Appeal’s ruling was then appealed to the House of Lords resulting in a decision which reviewed the following important aspects of patent law:

In contrast to the invention described in the Amgen claims, the TKT method did not involve any isolated (‘exogenous’) epo DNA. Instead, the TKT system ‘switched on’ epo DNA which was already present (‘endogenous’) in the cells in order to produce epo. TKT’s epo was produced using a method which did not literally infringe the Amgen patent⁴ but was argued by Amgen to infringe on a non-literal basis under the English doctrine of purposive construction. TKT’s ‘endogenous’ technique was quite different to the techniques which were known or would have been thought to work at the time of the patent. The known techniques involved taking ‘exogenous’ DNA encoding epo coupled to suitable regulatory genes, and inserting the whole lot into a host cell so that the cell would produce epo.

Faced with the question of whether the claims could be construed to cover TKT’s gene activation process, the House of Lords took the opportunity to review the English law on purposive infringement. Lord Hoffmann, with whom the other Lords agreed, gave the leading judgment. His starting point was Article 69 of the European Patent Convention (EPC) which states:-

He outlined the history of the Protocol on the Interpretation of Article 69 (the Protocol), and how it was born out of a compromise between the contracting states of the EPC some of which favoured the narrower English claim-based approach whilst others preferred the more generous approach of looking at the ‘essence of the invention’. The key aspect of the Protocol, Lord Hoffmann suggested, is striking a balance between fair protection for the patentee and reasonable certainty for third parties, and he went on to consider what this meant. He disagreed with Jacob LJ’s suggestion in the Rockwater case⁵ that fairness to the patentee meant that one must use the "widest purpose consistent with his teaching". He considered that if Jacob LJ had been suggesting that this meant giving the claims their widest possible meaning, that was not the correct approach. There was no presumption on the width of a claim and a patent might for one reason or another claim less than it teaches or enables.

Lord Hoffmann also rejected the US doctrine of equivalents as a means of providing fair protection to the patentee, because the Protocol had shut the door on any doctrine which extends protection outside the claims. He noted that the US Supreme Court had itself expressed some anxiety that the doctrine had "taken on a life of its own unbounded by the claims" and that the American Courts had sought to restrict the doctrine by applying so called ‘file wrapper’ or ‘prosecution history’ estoppel. The doctrine of file wrapper estoppel, as applied in the US, leads to the result that equivalence cannot be claimed for limiting integers that have been included in claims by amendment during prosecution. Combined with Lord Hoffmann’s earlier comments that the Courts of the UK, The Netherlands and Germany certainly discourage, if they do not actually prohibit, use of the prosecution file as an aid to patent construction, this suggests strongly that neither the doctrine of equivalents nor file wrapper estoppel will be applied in the UK in the foreseeable future.

Lord Hoffmann then reviewed the principle of construction currently applied by the English courts from Catnic, and concluded that they were "precisely in accordance with the Protocol". He suggested that the Protocol is intended to give the patentee the monopoly which a reasonable person skilled in the art, reading the claims in context, would think the patentee was intending to claim. Whilst Article 69 prevents equivalence extending protection outside the claims, Lord Hoffman made clear that it does not prevent equivalence being a part of the background of the facts known to the skilled person which would affect his understanding of the claims.

Significantly, Lord Hoffmann then went on to clarify the status of the so called ‘Improver’ or ‘Protocol’ questions which he had enunciated over a decade ago when sitting as the judge of first instance in Improver⁶. He suggested that it was important to distinguish them from the Catnic principle which is the "the bedrock of patent construction, universally applicable". The Protocol questions are, he said, only guidelines for applying that principle to equivalents that are more useful in some cases than in others. Amgen was a case where the guidelines were not useful, so rather than rigidly applying the Protocol questions, he went on to consider the facts of the case.

The question was not the meaning of the words used in the claim, but rather whether the invention operated at a level of generality that was such that it was irrelevant whether the epo DNA was endogenous or exogenous. This Lord Hoffmann said was a difficult question to put through the "mangle" of the Protocol questions because the answer is entirely dependent upon what the invention is. At first instance, Neuberger J had considered it to be the discovery of the epo DNA sequence, so that any method of using the sequence (be it exogenous or endogenous) could be covered. Conversely, the Court of Appeal thought the invention was the methods of making epo taught by the patent, limiting the monopoly to such methods. Lord Hoffmann, having noted that the discovery of the DNA sequence was not patenable and could not itself be the invention, agreed with the Court of Appeal. The methods in the patent were limited to exogenous DNA, and TKT’s process was therefore outside the claims.

Having resolved the question of infringement, Lord Hoffmann reiterated the fact that the Protocol questions are guidelines only. Although there had been historical differences, he considered that this approach was in line with the approach taken in Germany and The Netherlands. Furthermore, referring to the famous quintet of German cases, he suggested that the German courts now have their own guidelines for dealing with equivalents which resemble the Protocol questions, and that the German court has expressly stated that its approach is similar to Catnic.

The House of Lords’ acknowledgement that the Protocol questions are not the only way of answering the question of infringement throws open the possibility of other approaches. It is in relation to patents in rapidly developing, high technology fields such as the Amgen patent where the Court considered that the Protocol questions might have no useful function, and in today’s world there could be many such areas. Although TKT’s innovative technology was found on the facts to fall outside the patent’s claims, Lord Hoffmann indicated that there was no difficulty in principle with construing general terms in a claim to include embodiments which were not known when the claim was drafted. Such new technologies, he said, were another situation where the Protocol questions might be unhelpful.

The ruling, whilst it confirms that Catnic encapsulates the underlying principles of claim construction in English law and is in concert with our continental European neighbours, opens the door to tests other than Protocol questions being applied in high-tech cases.

Where a patentee has invented a new process for producing an old product, it has long been acceptable in the United Kingdom to claim products as made by his new process even though the products themselves were known. However, the position has for over 20 years been very different in many continental European jurisdictions. Neither our European neighbours nor the EPO accept the novelty of such claims⁸.

Throughout the litigation, TKT had argued that it was high time the UK changed its practice and started to follow the EPO’s approach. The relevance of the issue was that human urinary epo was a known product before the filing of the Amgen Patent. One of Amgen’s claims was for epo produced by a particular process. TKT’s argument was that such a claim to a known product made by a particular process should be anticipated by a disclosure of the product. This argument had been rejected both by Neuberger J and the Court of Appeal.

Lord Hoffmann explained that before the 1977 Act it was assumed that process claims could only be infringed by use of the claimed processes in the UK. This meant that claims to products made by a process were advantageous because, unlike the process claims, they could be used to restrain activities involving imported products made by the claimed process. He noted that the need for such additional protection was largely obviated by EPC Article 64(2), which was given effect in the UK by subsection 60(1)(c) of the 1977 Act. That subsection provides that the following acts constitute infringement in the UK:-

This means that acts pertaining to imported products can infringe so long as the products are obtained directly by the claimed process, even in the absence of any product claim⁹.

Turning to the prosecution of Amgen’s patent in the EPO, Lord Hoffmann noted that two claims had originally been framed as product by process claims. Epo is a protein which can have a variety of molecular weights, such that some epos weigh more than others. Faced with objections during opposition, Amgen amended one of the claims (claim 19) so that it set a test for infringement requiring the epo in question to be heavier than extracted urinary epo in order to infringe. Hence, by definition, the claim could not be anticipated by prior extracted urinary epo. The other claim (claim 26) survived opposition and, was granted covering epo products made by the process of claim 1. Though the granted claim was in effect a product by process claim, it was permitted by the EPO on the basis that the epo so produced was somehow different to the extracted urinary epo and could therefore be considered new. Neuberger J had found on the facts that there was no such difference, and Lord Hoffmann concluded that if the EPO had made the same finding, it would have held claim 26 anticipated.

Bringing this aspect of the English law on novelty into line with our European neighbours, he stated:-

TKT had suggested that both claims 19 and 26 of the patent were insufficient. The attack on claim 19 was based on the fact that the molecular weight of extracted urinary epo varied considerably so that the test for infringement set by the claim (i.e. i.e. being heavier than extracted urinary epo) was either impossible to carry out, because it required testing against all urinary epos, or was a lottery, because the finding of infringement depended on the particular urinary epo used.

Three separate arguments were levelled at claim 26. The first was a squeeze with infringement – if the claim covered epo made by TKT’s gene activation process then the patent was insufficient as it contained no teaching on the use of such processes (indeed, as noted above, they were thought not to work when the patent was drafted). The other arguments were based on the level of generality of the claim, which covered various analogues of epo (i.e. proteins expressed from different epo-like gene sequences), and epo analogues expressed by any host cell.

Considering the law on sufficiency, Lord Hoffmann, citing his earlier decision, Biogen¹⁰, confirmed that subsection 72(1)(c) of the 1977 Act requires that the disclosure enable the invention to be performed across the full extent of the monopoly claimed. He noted this is highly sensitive to the nature of the invention, and that the first step is to identify the invention and decide what it enables the skilled man to do. He had already done this for infringement – the invention was a process of making epo, and not epo itself or its properties, which were already known, or the epo gene sequence which in itself was only a non-patentable discovery.

With this in mind Lord Hoffmann assessed each of TKT’s arguments starting with the infringement squeeze. As this argument had been rendered redundant by the finding on construction, Lord Hoffmann did not express a concluded view. However, he took the opportunity to clarify the significance of a patent disclosing a ‘principle of general application’. In the House of Lords, Amgen had argued that its patent disclosed such a principle, allowing its claims to be framed in general terms without the need to disclose every embodiment.

Lord Hoffmann noted that Neuberger J, influenced by his view that the epo gene was the invention, had accepted that such a principle was disclosed by the disclosure of the epo gene sequence. He made clear that there is no magic to the concept of a principle of general application:-

He also noted that Neuberger J seemed to have thought that an invention was enabled if it disclosed something necessary to work the invention. This however was not enough. As well as disclosing the necessities the disclosure must be sufficient.

Although TKT had not argued that their process being inventive rendered claim 26 insufficient, Lord Hoffmann considered the implications for sufficiency of claims covering inventive improvements:-

This, in combination with the House of Lords’ assessment of the law on claim construction, leaves open the possibility of valid claims covering new, emerging technologies not contemplated at the time of a patent. Its application can also be seen in Lord Hoffmann’s analysis of the host cells argument. Like the Court of Appeal, he considered that, because the invention did not promise high levels of epo expression, the discovery of another cell which enabled high-level expression would be an improvement which did not have to be enabled by the specification.

On the claim 19 infringement test, Neuberger J had conclude that the lack of clarity in the specification made the claim insufficient. The test was in his opinion not just unworkable at the fuzzy edges of the claim, but made it impossible to determine whether any epo product infringed. The Court of Appeal had disagreed however, holding the claim valid. Lord Hoffmann noted that the Court had put a great deal of emphasis on the fact that the skilled person would be "seeking success". Though ordinarily correct – the skilled person would normally seek to make the invention work – seeking success did not seem to apply in this context:-

This approach has much to commend it. Where a claim sets a test which has nothing to do with working the invention but is simply a test for determining infringement, the test must be enabled at least to the extent that it can be operated by the skilled reader in a way that answers the question as to whether the claim is infringed. Having found claim 19 insufficient and 26 anticipated, Lord Hoffmann did not express a concluded view on the analogues argument.

The Amgen decision has clarified some of the fundamental principles English of patent law, and particularly how they apply to emerging technologies. It has also, by assessing construction and concluding we are in conformity and by changing the English courts’ approach to product by process claims, brought our patent law closer to that applied by our European neighbours. Patentees will welcome this increased uniformity, and Lord Hoffmann’s acknowledgement that valid patent claims can cover new innovative technologies will also be welcomed. However, whereas previously they could reasonably expect the scope of their claims to be determined by the Protocol questions, this may no longer be the case. They may therefore feel a little less certain about the exact scope of their monopolies.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.