Cambridge Antibody Technology v Abbott Biotechnology Ltd & Anor

The claimant in this action is Cambridge Antibody Technology Ltd (“CAT”). It is a research based company which is involved in the development and licensing of technology relating, among other things, to the production of antibodies. The first defendant is Abbott Biotechnology Limited and the second is Abbott GmbH & Co KG. I shall refer to them together as “Abbott”. The second defendant was known, until its acquisition by the Abbott group of companies in 2000/2001, as Knoll AG (“Knoll”). That company was part of the well known large pharmaceutical group, BASF. The first defendant is the second defendant’s holding company. In the course of this judgment it will be necessary to refer to one other company which was part of the same corporate group as Knoll. I will draw no distinction between them and Knoll save and to the extent that the context demands. Before me Mr Geoffrey Vos QC, Mr David Kitchin QC, Mr John Nicholls and Mr Richard Meade appears for CAT and Lord Grabiner QC, Mr Simon Thorley QC, Mr Marcus Smith and Mr Justin Turner appears for Abbott.

This action arises out of a dispute as to the meaning of the royalty provisions in two agreements entered into between CAT and Knoll under which the latter obtained a licence to use ground-breaking technology invented by CAT which was, at the time when the agreements were entered into, the subject of two patent applications. The first agreement was entered into in 1993 (“the 1993 Agreement”). It was replaced by another one entered into in 1995 (“the 1995 Agreement”). There is little difference in the language of the two and the parties agree that they were intended to have and did have, in substance, the same meaning and effect. I shall refer to them together as the Agreements. In what follows, I will refer to the Articles in the two Agreements, where appropriate, by putting the number of the Article against the year of the Agreement. Thus Article 12.01 of the 1993 Agreement would be referred to as 12.01/1993.

Knoll wanted to use the CAT technology, a brief description of which I will give below, so that it could develop certain pharmaceutical preparations. In particular in the early 1990’s it wanted to develop an antibody to a protein known as TNF. TNF is thought to play a major role in the chronic disease, rheumatoid arthritis. An antibody which targeted and neutralised that protein was considered a promising possible treatment of the disease. So it has proved. The result of the collaboration between CAT and Knoll/Abbott has been the production and marketing by the latter of a product called HUMIRA. It contains an antibody to TNF made by use of, inter alia, CAT’s technology. It has been described in the course of these proceedings and elsewhere as a potential block-buster drug. In the very short time since it has been placed on the market, its sales have grown fast and they are expected to reach a very high level. CAT’s royalties are set on a sliding percentage of the income generated by sales of HUMIRA.

In a nutshell, Abbott argues that the 1995 Agreement entitles it to offset against what is due to CAT 50% of royalties it pays to third parties in respect of other patented technology it uses or has used in the development of HUMIRA. This is subject to a contractual minimum according to which CAT will always receive a royalty of 2%. Because it has taken licences under a number of third party patents, Abbott says that the offset has reduced the amount payable to CAT to that minimum. Its royalty payments to CAT have been calculated accordingly. Further it intends to pay the same 2% royalty to CAT in respect of sales of further pharmaceutical preparations which have been or will be made using the technology licensed to it by CAT. It was the decision to develop an additional product which resulted in the 1993 Agreement (which had been drafted in contemplation of the development of HUMIRA) being replaced by the 1995 Agreement. A second pharmaceutical developed by Abbott by use, amongst other things, of the CAT technology is currently undergoing clinical trials. A third licence was entered into in 1999 which relates to a number of other possible targets for commercial development by Abbott. It is not central to this dispute and can be ignored.

CAT accepts that there is an offset provision in the Agreements. However it argues that it applies only to royalties which Abbott needs to pay to third parties in respect of the use of the technology licensed by CAT. Since all the licences relied on by Abbott are in respect of patents covering parts of the process for producing HUMIRA other than that involving CAT’s technology, the offset provision is not triggered. In the result Abbott should be paying a royalty for the bulk of its sales at the rate of about 5%.

CAT’s primary submission is that its construction of the 1995 Agreement is correct. But it has fall back positions. It says that if it is wrong on construction, the 1995 Agreement (which is the one now in force) should be rectified either on the basis of common mistake or unilateral mistake. Further it argues that Abbott is estopped from asserting its construction. All of these points will have to be considered in due course. However before turning to the Agreements and the factual context in which they must be construed, it is necessary to understand, at a fairly superficial level, the technology involved in making a product like HUMIRA. HUMIRA contains, as its active ingredient, an antibody made by using recombinant DNA technology; so-called genetic engineering. It is therefore necessary to know something of how antibodies work and something of how recombinant DNA technology can be harnessed to produce proteins of choice.

Antibodies

Antibodies are molecules which are generated by an animal’s immune system to assist in neutralising or destroying foreign matter, for example bacteria and viruses, which may have entered or be trying to enter the body. The surface of bacteria and viruses contain proteins. Proteins are very long molecules made up of building blocks, known as amino acids. It is the precise sequence and identity of the amino acids in the protein which, to a large extent, determine its size, shape, and chemical and physical behaviour. When a foreign protein enters, say, the blood stream of a human, the immune system recognises it as foreign and will set about trying to destroy or neutralise it. Each protein will have a number of places on its surface which are recognised as foreign and which can provoke an immune response from the host. These places are called antigens. An antibody is a molecule which the host immune system designs to lock onto a specific part of an antigen (called an epitope) on a foreign molecule. Since each antigen and epitope is different, different antibodies have to be made to lock on to each of them. Once an antibody has been built and it has attached to the antigen, it may interrupt some adverse behaviour of the protein on which the antigen is located, so that that behaviour is neutralised, or it may make the protein recognisable by the molecule-destroying systems in the host, with the result that the protein is either excreted by the host or destroyed.

Antibodies are themselves proteins. They are manufactured in specialist cells called lymphocytes. An individual lymphocyte cell can only produce a single design of antibody. If, therefore, the host wants to make five different antibodies to combat a foreign protein, it will be necessary to teach some lymphocytes to make one, others to make the second and so on. In fact there will usually be a number of different antibodies which can attach to a single epitope. Some will attach faster than others and some will have greater neutralising power than others. Once again each antibody molecule will be produced by a single lymphocyte. Since there are a vast number of possible permutations of foreign proteins and their antigens which the host has to be capable of responding to, each lymphocyte has to be capable of making any one of a vast number of bespoke antibodies. For the purpose of these proceedings it is not necessary to understand how lymphocytes do this. Suffice to say that all lymphocytes are capable of being programmed to make one of a vast array of possible antibodies.

Because they are proteins, antibodies are made by the protein-manufacturing apparatus within the lymphocyte cells. But because they are proteins they can also generate an immune response if they are put into an alien immune system. Thus an antibody generated in a mouse (called a “murine” antibody), if injected into, say, a human will generate an immune response in the human. This, as we shall see, is important when it comes to designing artificial antibodies for use as therapeutic agents in humans.

Antibodies are a group of proteins (strictly glycoproteins) known as immunoglobulins (“Ig”). There are a number of classes of Igs. For present purposes the difference between them is not important. This case is particularly concerned with a class known as immunoglobulin-G (“IgG”). Understanding something about the structure of these is useful as a background to some of the disputes which have arisen between the parties.

An IgG is a protein which is made up of four chains of amino-acids which chains are held together by chemical bonds. There are two identical long chains, referred to as the Heavy Chains or Hc, and two identical short chains, referred to as the Light Chains or Lc. They are held together to create a symmetrical Y-shaped molecule which can be illustrated diagrammatically as follows:

The Heavy Chain can be regarded as being made up of four distinct domains linked together and the Light Chain can be regarded as being made up of two domains. These are depicted diagrammatically in the above illustration. In each chain, one of the domains is said to be variable, the remaining chains are said to be constant. The variable domains are those located at the end of the two arms of the Y-shaped molecule. They are frequently referred to as the Hv (i.e. Heavy Variable) and Lv (i.e. Light Variable) domains. The domains are referred to as variable and constant to reflect the extent to which the amino-acid sequence in them varies from antibody molecule to antibody molecule within the same animal. Thus, for example, a human being will have a large number of antibodies in his bloodstream, depending on the foreign materials (e.g. bacteria and viruses) which he has had to raise immune defences to. This will include many IgG molecules. However the constant regions in one such molecule will be substantially identical to the constant regions in all the other IgG molecules his lymphocytes produce. But the variable regions of the antibodies made to defend against, say, whooping cough will differ from the variable regions of the antibodies made to defend against, say, mumps. Indeed, because most foreign proteins have more than one antigen and the body’s defence mechanism will make antibodies against most of them, and because that mechanism is likely to make a number of different antibodies against each antigen (having different abilities to lock on and neutralise), there will be a number of antibodies for each foreign proteins and each of them will have variable domains which differ from the variable domains of the other antibodies. It is the ability of an immune system to make variations in the variable domains which contributes to its ability to create bespoke antibodies which lock onto single epitopes. This ability is further enhanced by the fact that at the very outer end of each variable domain there is a hyper-variable region made up of three small lengths of amino acids. These are called complementarity-determining regions (CDRs). As their name suggests, the amino acid sequence within these regions is highly variable when compared with the rest of the antibody molecule. These CDRs dictate the precise antigen-binding characteristics of the antibody. The CDRs on the Hv and Lv domains co-operate together to produce such characteristics.

It is possible to cut an IgG molecule into pieces or to manufacture pieces of it. For example, it is possible to isolate each of the two arms of the Y-shaped molecule. These are called Fab (i.e. Fragment Anti Body) fragments. They will contain the CDRs, the Variable Domains and some of the Constant Domains. On the other hand it is possible to isolate the Variable Domains alone (containing their CDRs). If the Variable Domains at the end of one arm of the Y-shaped molecule are isolated (i.e. the Hv and Lv domains), these two pieces can be bound together with a short chain of amino acid (called a peptide linker) to produce something which is called a single-chain Fv fragment (i.e. single chain Fragment Variable). This is abbreviated for simplicity to scFv. All of these concepts are illustrated in the following drawing.

Simple concepts in genetic engineering

The field of genetic engineering has developed spectacularly over the last three decades. It has involved the development and use of many sophisticated techniques, much of it involving manipulation of individual molecules, and the design, manufacture and use of specialist equipment and reagents. What is set out below cannot begin to give a feel for the breadth and complexity of what is involved. However, for the purpose of understanding the issues in this case, detailed knowledge of the technology is not required.

Proteins are major building blocks in living organisms. As mentioned above, each protein is made up of a specific sequence of amino acids. A protein may consist of a chain containing hundreds of such amino acids. The precise sequence of amino acids and the shape of the complete molecule are major determinants of the physical and chemical behaviour of the protein. The sequence in which specific amino acids are joined together to make a specific protein is determined by a living cell’s genes. Each gene will contain, in code, the different sequences of all the proteins which that cell is likely to need to use during its life. In mammals, that code is to be found in the DNA in the genes contained within its cells. If one wants to make a protein artificially, it would be possible, in theory at least, to add one amino acid to another in a defined sequence in a reaction flask. This would be slow, difficult and would give a low yield. However all living cells contain protein-generating equipment. They must do so otherwise they could not produce the proteins which they need to live and function. At its most basic level, genetic engineering uses this existing equipment to make proteins, or fragments of protein, to order. In essence a piece of DNA which contains the code for a particular protein is inserted into a host cell in such a way that the host cell treats it as its own. The host cell then uses it internal machinery to produce (or “express”) the protein of choice.

This process involves a number of stages, each of which involves the use of complex science. First, it is necessary to identify the particular protein or fragment which is to be produced. Second, it is necessary to find or reverse engineer the particular piece of DNA which contains the code for it. Third, it is necessary to cut out or manufacture that piece of DNA and make multiple copies of it. Fourth, a suitable host cell into which the DNA can be inserted must be selected. For example, in some cases it is possible to insert a DNA sequence in specialised yeast cells. Industrial fermentation of the yeast will produce the protein or fragment of choice, rather than alcohol. Other possible host cells include “immortal” mammalian cells, such as cell lines derived from certain cancers. Fifth, a method must be found for inserting the DNA into a vehicle (or “vector”) which is able to carry the DNA sequence into the host cell and insert it in the correct location. This means that it must be placed in a location which will control its accurate expression. Sixth, the host cell will need to be designed or selected to ensure that the desired protein is expressed in quantities which are large enough to meet the requirements of the program in which the genetic engineering is being used. Seventh, it will be necessary to have in place suitable equipment and processes to purify the protein or fragment. Needless to say, each of these steps involves a number of sub-steps, each of which may be complex and time-consuming.

The application of genetic engineering technology to the production of antibodies

Over many years attempts have been made to harness genetic engineering technology to the production of antibodies which could be used in the therapeutic treatment of humans. In theory it should be possible to make large quantities of pure antibodies which could be given to patients to boost or supplement their own immune systems. However there are a number of problems involved in doing this. For present purposes one of the most significant is the fact that antibodies, being proteins, may themselves generate an immune response from the immune system of the patient into whom they are injected. If that happens, the artificial antibodies will be broken down or neutralised. The faster and more completely this happens, the less effective the antibodies will be as a therapeutic treatment. On the other hand, humans do not normally raise an immune response to their own antibodies. Such proteins are recognised as “friendly” and are not attacked. For that reason, as Mr Simon Thorley QC puts it, the holy grail in this area was to produce in large quantities and by genetic engineering antibodies which the patient’s immune system would accept as completely “human”.

It is possible to make antibodies to a particular foreign protein by injecting that protein into an animal, waiting for the latter’s immune system to react and then harvesting the antibodies it produces. Those antibodies can then be selected out and purified. In this process, the animal is used as a factory which can be made to produce an antibody of choice. By injecting the foreign protein or agent (e.g. a pathogenic bacterium) into a sufficient number of mammals it is possible to harvest significant quantities of antibodies. Since such animal antibodies are designed to lock onto the epitopes on the foreign protein, they should be capable of doing the same thing if they are injected into a human. It is possible to harness genetic engineering techniques to manufacture such antibodies. For example, once a specific murine (i.e. mouse) antibody has been isolated, it is possible to find or work out the DNA which codes for it. That DNA, if used in the way indicated above, should be capable of being inserted into a host cells (for example immortal – i.e. cancer – cells from Chinese Hamster Ovaries, called CHO cells) and used to produce large quantities of murine antibodies. This has been done. Unfortunately such an antibody is 100% murine. It is therefore likely to induce an immune response if injected into a human. This is called a HAMA (Human Anti Mouse Antibody) response. The patient’s immune system reacts to the injected artificial antibody and, once its own immune system has learned to identify and react to this foreign protein, eradicates it. Such artificial antibodies can therefore only be used, if capable of being used at all, for a very short period of treatment.

The obvious solution to this would be to produce and harvest human antibodies. For obvious ethical reasons, it is not normally possible to use humans in a similar way to laboratory animals for the purpose of harvesting antibodies on demand. So some way had to be found to make and harvest antibodies which were less non-human. Thus it was desired to reduce the amount of murine protein in the antibody. This has been done by generating murine antibodies, chopping off the Heavy Chain and Light Chain variable regions (i.e. the parts which are critical to binding the antibody to the target antigen) and grafting them onto the constant region of human antibodies (which are well known). Such antibodies are called chimaeric. Once a chimaeric antibody has been made, it is possible to make a DNA sequence which codes for that protein and it is possible to insert it into host cells for expression. Since the antigen binding sites come from the murine antibody which tackled the foreign target (e.g. bacterial proteins), the chimaeric antibody should have an antibody function which is similar to the full murine antibody, but, because most of it is human, it should produce a lesser HAMA response. However, there is still much of the mouse protein left in the molecule so some HAMA response is likely. The next step is to cut out from the murine antibody only the CDRs and to graft them onto human antibodies. This too has been done, but there is still a small amount, perhaps 5%, of murine amino acid sequences present in the final antibody. Some HAMA reaction is still to be expected. It will be understood that any system of making antibodies by genetic engineering which involves copying any part of the final product from a mouse (or other laboratory animal) antibody, will incur the risk of a HAMA-type response.

The different types of possible genetically engineered antibodies are illustrated below.

It follows that the ultimate goal is to make an artificial antibody which is entirely human and none of which is copied from an antibody derived from a laboratory animal. That is what the technology licensed by CAT to Abbott makes possible.

The technology licensed by CAT

The technology licensed by CAT consists of two parts which co-operate together to put you well on the way to producing a fully human antibody. The two parts are now covered by patents. The first is referred to as Winter 2, the second as McCafferty, after the names of one of their respective inventors.

As explained above, the immune system of a human (or any other mammal) has to be capable of making antibodies to react to and incapacitate a vast range of different foreign proteins. Its genes are designed in such a way that it can do this. How it does so is not necessary to explain. What Winter 2 does is to mimic in some ways how the human immune system works. Essentially it discloses and monopolises a way in which to generate vast numbers of variants of the variable domains for human antibodies. If sufficient such domains are generated they will include all the variants which a human immune system is capable of producing in order to meet a challenge from a foreign protein. The Winter 2 patent teaches how to produce rapidly tens of millions of DNA sequences coding for such human variable domains. Unfortunately its success is also its weakness. It produces a “haystack” of variable domain sequences. What the would-be producer of a therapeutic antibody treatment wants is the variable domains which, when attached to the rest of a human antibody molecule, will work best to knock out the target foreign protein. That is to say he needs to find the “needle”. It should not be thought that there will be only one “needle” for a particular epitope. It is likely that a number of variable domains will be able to bind to each epitope. Some will be better than others. The library of sequences produced by Winter 2 will contain some which have high affinity (i.e. will bind strongly and quickly to the epitope), some which have high specificity (i.e. will bind only one target epitope and none other) and some which have both high affinity and high specificity. The ideal is to find the latter.

This is where McCafferty comes in. Essentially it involves a method by which each of the sequences in a DNA library (such as, but not limited to, those created by use of the Winter 2 technology) is loaded onto a bacteriophage (or simply “phage”). A phage is a small particle which acts like a virus towards bacteria. It can invade bacteria and multiply inside. By the McCafferty process, not only are the individual DNA sequences loaded onto phages but the phages are capable of expressing (i.e. producing) the protein fragment coded for by each inserted sequence. Those expressed protein fragments are carried on the surface of the phage. McCafferty teaches how they can be searched through quickly so as to identify the ones which have the best DNA sequences to produce a particular antibody. The detail of how this is done is not central to this dispute, but in essence it works as follows. A surface of, say, a plate has the target antigen fixed to it. Vast numbers of bacteria which have been infected by the modified the phages are allowed to run over the surface. The very small number of bacteria which contain phages which carry the DNA sequence for effective variable regions for antibodies to the antigen will have those antibody regions on their surface. They will therefore stick to the antigens which are stuck to the surface of the plate. All the millions or tens of millions of other bacteria will not stick and can be washed off. This results in the DNA sequences of choice being retained. The process is called “panning”. The process is not limited to a single sweep through the library. The first sweep may well turn up some promising candidates which are then used to refine the search on a second sweep through. If all goes well, this iterative process will enable the user to select out variable domains for both the Heavy and Light Chains which together will form the “business end” of a complete antibody which is good both at locking exclusively to the target foreign protein and at helping to neutralise or destroy it. The idea is to find the Heavy Chain and Light Chain scFvs which produce the best pairing.

It is possible to us scFvs as antibodies in their own right since they will bind to their equivalent epitopes. However they are much smaller than full antibodies, so they are likely to be destroyed much more quickly. They also contain the polypeptide linker between the two variable regions (see the illustration in paragraph 13 above). This is “foreign” so that it may produce an immune response. Furthermore, they lack part of the structure of the full antibodies and, in some respects, the rest of the structure can help the antibody function. Nevertheless, because they can act as antibodies in their own right, it is possible to carry out the iterative process described in the previous paragraph without building full antibodies at each stage. In other words fragments produced in the first run through the phage display process could be tested for affinity and specificity with the best ones being selected to be used as the starting point for the second run. It is not necessary, after each run, to take the promising fragments and build them into full antibodies before the next sweep is undertaken. This was explained by Professor John Isaacs, CAT’s expert. Nevertheless building and expressing full antibodies can be done at the end of each sweep. This may give a better indication of which scFvs are most promising. It was what was done by Knoll in the development stages which led, eventually, to HUMIRA.

In summary, Winter 2 enables one to make a vast “library” of potentially useful human DNA sequences. McCafferty deploys its “phage display” techniques to select out from such a library those sequences which are most likely to contribute to the construction of a highly effective human antibody. What is produced at the end of these two processes are fragments, not whole antibodies, but they are the crucial working end of antibodies. They can be added to the rest of a standard human antibody molecule to create a full human antibody. It is this patented library and phage display technology which was licensed to Abbott. It has been described as a breakthrough. It was used to build the DNA sequence which, when put through all the other stages referred to above, allowed the expression of commercial quantities of human antibodies to TNF. Those antibodies were then put together with other materials and placed in a special syringe. That is the finished product which is sold under the name HUMIRA.

Other technology relevant to this dispute

The whole field of genetic engineering has been the beneficiary of hectic research. The result is that there is much new technology and numerous patents in this field. Making a commercial product like HUMIRA, as indicated above, is a lengthy, multi-stage process involving sophisticated ingredients, methods and apparatus, some of which may be patented. In this case reference has been made to a number of third party patents. Some of them also cover the same technical areas of library creation and phage display as Winter 2 and McCafferty, that is to say operation of the technology in those patents may involve the utilisation of technology patented by other parties. As will be explained below, a number of such third party patents exist. However, CAT’s library creation and phage display technology is only a part of the total process which has lead to the design, manufacture and sale of HUMIRA. Other parts of the process may fall within the scope of third party patents. A number of such patents have been mentioned during the course of this trial. They have been referred to at times as being in respect of “post-CAT” technology. That is a useful shorthand which I adopt in this judgment. It is, however, no more than a convenient label.

The approach to construction

As I understand it, there is no serious dispute between the parties as to what is the correct approach to construing a contract. That has been the subject of recent consideration at a high level. What one is trying to do is to discover, from an objective assessment of the words used, the intention of the parties to the agreement. The parties are assumed to have recorded their true intent in the words they chose. It is not permissible to have regard to the subjective intent of one or other of the parties. This means that evidence as to the state of mind of the parties during the process of negotiation and documents produced for the purpose or recording the negotiations are not admissible. This can cause problems where, as here, one or more parties seeks rectification. In relation to the latter (and subject to a legal argument of Lord Grabiner’s which I will consider below), the intention of the parties and documents evidencing it are not only admissible, they are central. Where simple construction and rectification are run side by side, it is necessary to bear in mind this difference and, when construing the document, to exclude from consideration all the material which is only admissible in relation to rectification.

Mr Vos relies on one of the most recent authorities on construction, Investors’ Compensation Scheme v. West Bromwich Building Society [1998] 1 W.L.R. 896. He places particular reliance on the following passage in the speech of Lord Hoffmann:

“(1)

Interpretation is the ascertainment of the meaning which the document would convey to a reasonable person having all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contract.

(2)

The background was famously referred to by Lord Wilberforce as the “matrix of fact”, but this phrase is, if anything, an understated description of what the background may include. Subject to the requirement that it should have been reasonably available to the parties and to the exception to be mentioned next, it includes absolutely anything which would have affected the way in which the language of the document would have been understood by a reasonable man.

(3)

The law excludes from the admissible background the previous negotiations of the parties and their declarations of subjective intent. They are admissible only in an action for rectification. The law makes this distinction for reasons of practical policy and, in this respect only, legal interpretation differs from the way we would interpret utterances in ordinary life. The boundaries of this exception are in some respects unclear. But this is not the occasion on which to explore them.

(4)

The meaning which a document (or any other utterance) would convey to a reasonable man is not the same thing as the meaning of its words. The meaning of words is a matter of dictionaries and grammars; the meaning of the document is what the parties using those words against the relevant background would reasonably have been understood to mean. The background may not merely enable the reasonable man to choose between the possible meanings of words which are ambiguous but even (as occasionally happen in ordinary life) to conclude that the parties must, for whatever reason, have used the wrong words or syntax: see Mannai Investments Co Ltd v. Eagle Star Life Assurance Co Ltd [1997] AC 749.

(5)

The “rule” that words should be given their “natural and ordinary meaning” reflects the common sense proposition that we do not easily accept that people have made linguistic mistakes, particularly in formal documents. On the other hand, if one would nevertheless conclude from the background that something must have gone wrong with the language, the law does not require judges to attribute to the parties an intention which they plainly could not have had. Lord Diplock made this point more vigorously when he said in Antaios Compania Naviera S.A. v. Salen Rederierna A.B. [1985] AC 191, 201: ‘if detailed semantic and syntactical analysis of words in a commercial contract is going to lead to a conclusion that flouts business commonsense, it must be made to yield to business commonsense.’” (p 912H-913E)

In BCCI v. Ali [2002] 1 A.C. 251, Lord Hoffmann explained at paragraph 39 what he meant at (2) above:

“The background is however very important. I should in passing say that when, in Investors Compensation Scheme Ltd v. West Bromwich Building Society [1998] 1 WLR 896, 913, I said that the admissible background included “absolutely anything which would have affected the way in which the language of the document would have been understood by a reasonable man”, I did not think it necessary to emphasise that I meant anything which a reasonable man would have regarded as relevant. I was merely saying that there is no conceptual limit to what can be regarded as background. It is not, for example, confined to the factual background but can include the state of the law (as in cases in which one takes into account that the parties are unlikely to have intended to agree to something unlawful or legally ineffective) or proved common assumptions which were in fact quite mistaken. But the primary source for understanding what the parties meant is their language interpreted in accordance with conventional usage: “we do not easily accept that people have made linguistic mistakes, particularly in formal documents”. I was certainly not encouraging a trawl through “background” which could not have made a reasonable person think that the parties must have departed from conventional usage.”

In BCCI v. Ali, Lord Nicholls summarised the principles at paragraph 26:-

“The meaning to be given to the words used in a contract is the meaning which ought reasonably to be ascribed to those words having due regard to the purpose of the contract and the circumstances in which the contract is made”.

Mr Vos also submits that evidence of the parties’ subjective intentions is admissible to show that the parties negotiated on an agreed basis that the words used bore a particular meaning. For this he relies on Chitty on Contracts (29^th ed (2004) at paragraph 12-119) and the judgment of Kerr J in Partenreederei M.S. Karen Oltmann v. Scarsdale Shipping Co Ltd [1976] 2 Lloyd’s Rep. 708. The trouble with this is that, even if it represents the law, it leads one perilously close to looking generally at the subjective intention of the parties when construing a contract. On the current state of the law, that is not permissible. On the facts of this case, I do not think that it is necessary to go down this path.

The factual matrix

There is a large measure of agreement between the parties as to what is the relevant factual matrix against which the Agreements are to construed. As Lord Grabiner emphasises, it does not include the details of negotiations between the parties or their subjective intentions. Those are matters which are likely to be relevant, if at all, to issues of rectification and estoppel. Since those issues are raised in this action, it is inevitable that the evidence before me has extended far beyond what would be admissible on the issue of construction alone. In such circumstances, as noted above, it is particularly important to keep in mind the limited nature of the evidence which is admissible on the issue construction. I have attempted to do so in this judgment.

I set out below the relevant factual matrix. Much of this is derived from the Skeleton Arguments used by Lord Grabiner and Mr Vos. In the few areas where there is a dispute between the parties as to what qualifies as admissible factual matrix, what is set out below consists of my determination of how such disputes are resolved. The parties split the factual matrix into a number of headings. I will do likewise.

CAT’s technology, and the history of its patents.

Applications for patents for the library and phage display technology had been made before CAT and Knoll commenced negotiations. At the time of the 1993 Agreement, Winter 2 and McCafferty were pending patent applications. No patents had been granted. By time of the 1995 Agreement the first European Patent for Winter 2 had been granted. At that time, McCafferty was still a pending application. It was understood by the parties that the key aspect of the Winter 2 technology was its ability to generate extremely large libraries of different antibody fragments. The key aspect of McCafferty technology was phage display.

It is part of the factual matrix that collaboration was being considered whereby such technology would be used to produce human antibodies which neutralise certain human antigens, and in particular TNF, for therapeutic administration to patients. CAT had the power to grant licences under the Winter 2 and McCafferty applications in relation to the library and phage display technology and was willing to do so. CAT was not a pharmaceutical company. Although it was a young company with few employees, one of its strengths was its experience and knowledge of its library and phage display technology.

Knoll’s expertise and work relating to TNF .

Knoll was the pharmaceutical arm of BASF. As such it had the backing of one of the world’s largest chemical companies. Prior to its involvement with CAT, Knoll had been working on developing TNF. Knoll had been developing a mouse-derived (murine) antibody to TNF, known as MAK 195. Knoll was in a position to convert scFv fragments into antibodies using recombinant technology, for testing and thereafter intended to manufacture a suitable neutralising antibody in the United States for sale (subject to the successful outcome of clinical trials).

Technologies offering alternative routes to Knoll to an anti-TNF antibody.

There were alternative routes to making antibodies. Most of them resulted in the production of antibodies which were partly foreign, such as the chimeric antibodies discussed above.

The route offered to Knoll by CAT in isolating the variable domains and encoding genes: the purpose of the collaboration.

The purpose of the collaboration at all times was the eventual production and sale of human antibodies which neutralise human antigens (in particular TNF) for use in therapy. The purpose of the 1995 Agreement was to allow Knoll to develop antibodies to a new target, namely a protein known as IL-12, which plays a role in auto-immune diseases such as multiple sclerosis, psoriasis and Crohn’s disease. CAT carried out its work at its premises in Cambridge, England. Knoll would carry out its stages in development and manufacturing in the United States. It was developing a state of the art recombinant antibody manufacturing facility in Worcester, Massachusetts. It rather than CAT had expertise in expressing proteins.

The parties’ understanding of the patent environment prevailing at the time in relation to antibody gene library and phage display technology and in relation to technology subsequent to the phage display step.

The utilisation of CAT’s library and phage display technology and the development of antibodies to TNF would take place in an field in which there were many third party patents some of which might impact on one or more stages in the development, manufacture and sale of the finished product.

The parties’ understanding includes the contents of the so-called Patent Confidential document (referred to throughout the trial as the Pat Confid document) which was supplied by CAT to Knoll on or about 28 July 1992 and which gave an overview of the patent landscape at that time. The contents of that document will be considered in more detail later. This document identified third party patents which might cover or impinge upon (a) the antibody gene library technology of Winter 2, (b) the phage display technology of McCafferty and (c) post-CAT technology. For present purposes it should be noted that group (b) included a patent called Neuberger and group (c) included two patents referred to as Boss and Cabilly. The latter patents related to the expression stage, that is to say the stage at which the full antibody molecules are produced from host cells. Amongst other things, the Pat Confid document made it clear that there were “interference” proceedings on foot in the United States concerning the latter two patents. That means that the owners of Cabilly argued that their invention was, in substance, the same as the invention in Boss but of earlier date. Under American patent law it is not possible for two patents to cover identical inventions. The purpose of an interference is to determine whether the patents are for identical inventions and, if so, to determine which one was first. It is that one which survives. The patent for the later invention is invalidated in whole or in part.

The nature and purpose of the parties’ arrangements concerning the MRC Patents.

This is an issue which raised its head throughout the trial. It relates particularly to the Neuberger and Boss patents. As indicated in the previous paragraph, the Neuberger patent was of particular significance to CAT’s technology. In particular, in the form in which it was issued in Europe, it covered, or might cover, the phage display technology of McCafferty. If the phage display technology was to be used, a licence under Neuberger might well be required also. If the fragments to be produced by the phage display technology were used to make full length antibodies in host cells a licence under Boss (or Cabilly) might well be required. This is because the Boss patent describes the expression of both heavy and light recombinant immunoglobulin genes in a single host cell (yeast) and the recovery from the cell supernatant of a functional immunoglobulin (antibody) molecule. Its claims are essentially identical to claims filed in the Cabilly application which resulted in the interference proceedings in the US to which I have referred already. Boss (like Cabilly) does not relate to the process of identifying scFv fragments, or the identification of genes coding for variable regions of a particular specificity. It is a patent directed to the expression of antibodies from a single host cell once such genes have been isolated.

Boss and Neuberger were owned by a third party, Celltech. However both were licensed to the Medical Research Council (“MRC”) under circumstances which allowed it to grant sublicences. CAT was closely connected to the MRC and could be expected to secure such sublicences. However the terms upon which the MRC was allowed by Celltech to grant licences was restricted in that Boss could only be sub-licensed by the MRC in conjunction with Neuberger. In other words, the Boss patent would only be licensed to a sublicensee of the Neuberger patent who needed it in support use of the latter. In fact, if sub-licences under these two patents were to be secured, it was understood by the parties that it would have to take the form of a licence from the MRC to Knoll who would then have to pass on a sub-sub-licence in respect of Neuberger to CAT. This is because the terms of the licence between Celltech and the MRC were such that such sub-sub-licenses could be granted under Neuberger, but not under Boss. So, any MRC licence would have to go through Knoll to CAT, rather than through CAT to Knoll.

Thus, through the MRC, CAT could provide Knoll with a route to Boss. It was apparent that it would be very useful for Knoll to obtain a route through to Boss since Knoll was probably going to manufacture in the U.S. Boss was granted in the U.S. and covered manufacture of recombinant antibodies. The possible expression routes which Knoll was likely to use could well to fall within the scope of Boss.

The respective roles of CAT and Knoll in the collaboration and, in particular, in the development of HUMIRA.

As Lord Grabiner puts it, the respective roles of the parties were that CAT was going to produce fragments which would be converted to full antibodies by Knoll and tested by Knoll for their neutralising properties. Once a candidate antibody had been identified, Knoll would produce the antibody, characterise it, perform clinical trials and, if successful, sell a product containing such antibody as an active ingredient to the public.

The relevant terms of the 1993 Agreement

The 1993 Agreement starts with a Recital in the following terms:

“Knoll and CAT wish to enter into a collaborative development program aimed at isolating human neutralising antibodies to certain human antigens, beginning with TNF. Knoll is to take an exclusive, worldwide license to various patents and know-how in the Exclusive Field as defined below relating to phage-antibody technology. The collaboration and the license are to be on the terms and conditions set out in this Agreement.”

Certain terms in the Agreement are defined in Article 1/1993. They include the following which are material to this Recital:

“‘Exclusive Field’ means the isolation, development, manufacture, use and sale of the Antibodies and all antibodies (excluding mAb 32) and molecules derived using the technology covered by the Patents for human in vivo use.

‘The Antibodies’ mean antibody molecules or fragments or derivatives thereof, identified, produced or derived under the Research Program binding to The Antigen.

‘The Antigen’ means human tumour necrosis factor () (TNF )

‘Patents’ mean PCT applications WO 90/5144 (Winter 2) except as pertaining to single variable domains as described in claims no. 32 – 40, as published on May 17, 1990, and WO 92/1047 (McCafferty), the patents when granted, including any patents which derive from the Patents and any divisions, renewals, continuations, continuations-in-part, extensions or reissues.

‘Research Program’ means the research program aimed at the identification, production and development of The Antibodies to be conducted under Article 2 of this Agreement”

Thus the recital records that the parties are going to join in a development program. That will involve Knoll in taking a licence both to patents and know-how relating to “phage-antibody” technology. At the time the licence was entered into the patents were only at the application stage although it was expected that they would eventually mature into granted patents and might give rise to further derivative patents.

The nature of the Research Program is the subject of Article 2. Article 2.01/1993 states that the aim of the Program (which is to be funded by Knoll – Article 2.02/1993) is to achieve the goals set out in Schedule 1 to the Agreement. The latter splits the development into 8 “quarters” starting with fragments in the Winter 2 library and then going through to the Eighth Quarter which is:

“BBC [i.e. Knoll] to convert lead clones into full-length antibodies if not already. Optimisation of expression. Initiate phase I/II trials.”

Thus the research program goes right the way through to optimisation of expression, maximizing the production of the full-length antibodies, and clinical trials. These latter stages are needed to ensure that the full-length antibodies can be made in sufficient quantities to meet commercial demands for the finished pharmaceutical preparation and to test the latter for efficacy and lack of unacceptable side-effects prior to it being given regulatory approval.

Article 3/1993 contains provisions in accordance with which Knoll will fund CAT’s input into the development work. Redacted so as to leave out financial figures which are irrelevant to this dispute, it is in the following terms:

“3.01

Knoll will pay CAT a non-refundable fee to fund the Research Program of £xxx within thirty (30) days of the Effective Date.

3.02

In addition to the payment to be made pursuant to sub-clause 3.01 above, Knoll will pay CAT milestone payments within thirty (30) days of CAT achieving each of the CAT Technical Milestones 1 – 3 identified in Schedule II to this Agreement. Payments are to be made as follows: …

3.03

Knoll may, in addition to the payments to be made pursuant to sub-clause 3,01, 3.02 and 3.04 herein, make the following payments to CAT or be subject at CAT’s option to conversion of the exclusive license granted by this Agreement to a non-exclusive license, all other terms of this Agreement remeining (sic) the same:

(a)

… within 30 days of achievement of CAT Technical Milestone 4 or twelve (12) months after the achievement of CAT-Technical Milestone 3, whichever is earlier.

(b)

… within thirty (30) days of initiating the first human clinical studies for the first Product, or within 30 months of achievement of CAT Milestone 4, whichever is earlier; and

(c)

… within thirty (30) days of Knoll’s receipt of formal PLA approval for the first Product in any given country other than the United States of America, or within nine years of the IND filing in the United States for that Product, whichever is earlier.

3.04

Knoll will, in addition to the payments to be made pursuant to the foregoing, make the following payments to CAT:

(a)

… within thirty (30) days of the start of Phase III clinical studies of the first Product:

(b)

… within thirty (30) days of Knoll’s receipt of formal PLA approval in the United States of America for the first Product, said payment to be reimbursable as follows …”

It should be noticed that some of these payments are triggered by events late in the development program. See, for example, Articles 3.03(c)/1993 and 3.04(b)/1993 which refer to PLA approval. The latter is defined in Article 1.17:

“ ‘PLA’ means a Product Agreement Application, or other application approving the marketing of the Product, which is submitted to the FDA.”

The FDA is the Federal Drugs Administration in the United States of America which has the task of determining whether pharmaceutical preparations are to be licensed for sale and use in that country. Taken together, it will be seen that the various funding obligations on Knoll are triggered sequentially right the way through to the time when the Product is given regulatory approval to be marketed. The latter expression is also defined in Article 1.08/1993:

“ ‘Product’ means any pharmaceutical composition containing one or more of the Antibodies.”

What is meant by the word Product is also indicated by the definition of First Commercial Sale in Article 1.11/1993:

“ ‘First Commercial Sale’ means the first sale or other disposition for value of the Products, in a final dosage form packaged for the ultimate consumer, to an independent third party following regulatory approval. The term ‘First Commercial Sale’ shall not include sales and disposals for experimental purposes or for purposes of clinically or otherwise testing any Product.”

It seems to me to be clear that the Product is, or includes, the final packaged dosage form of the pharmaceutical preparation containing the full-length anti TNF antibodies to be derived from fragments isolated by the CAT library and phage display technology. It is this final form which is given clearance by the FDA. As will be seen below, this is also consistent with the last sentence of the definition of Know-how in Article 1.14/1993. The only reason for pointing this out is that the Agreement, somewhat confusingly, also refers occasionally to “Drug” which is defined in Article 1.18/1993 as meaning:

“a drug which is made, isolated and/or developed by Knoll and/or its Affiliates wholly or partly using The Antibodies or derived from The Antibodies”

Save that the latter also includes pharmaceutical preparations “derived from The Antibodies”, there appears to be no difference in scope between Product and Drug. No one suggests otherwise. This is also consistent with other provisions, such as the definition of “Net Sales” (see below) and Knoll’s warranty in Article 14.02(b)/1993;

“that it will use all commercially reasonable efforts to … develop saleable Products of good merchantable quality and to bring those Products to market within the time usually taken to bring products of a similar nature to market and to satisfy the market demand for those Products.”

The Product of the Agreement is the finished article which is sold into the market. It is this which has to be of good merchantable quality. This is the final packaged form of HUMIRA. It is on the net sales of the latter that Knoll has calculated the royalties due to CAT.

Recapping at this point, the various funding payments to CAT occur at stages from the very beginning to the very end of the process leading to the marketing of a commercial pharmaceutical product. This includes stage payments made in respect of parts of the development which are entirely the responsibility of Knoll, that is to say more or less everything after the phage display and selection stage in the program. They include, but are not limited to, stage payments to be paid on reaching each of four Technical Milestones. The latter are defined in Schedule II/1993 as follows:

“SCHEDULE 2^*

M1: isolation of ten or more “human antibody fragments: which bind specifically to human TNF i.e. do not bind with K_a’s  10⁵ M^-1 to a panel of ten other human antigens)

M2: Isolation of four or more “human antibody fragments” that bind specifically to human TNF with K_a’s  10⁷ M^-1 and neutralize 50% of TNF bioactivity in vitro with a MAb:TNF molar ratio of  100:1

M3: Isolation of two or more “human antibody fragments:” that specifically bind to human TNF with K_a’s  10⁸ M^-1 and neutralize 50% of the TNF bioactivity in vitro with MAb:TNF molar ration of 10:1

M4: Generation of two or more intact IIF human anti-TNF antibodies with K_a’s  10⁹ M^-1 and which neutralise 50% of TNF bioactivity in vitro with a MAb:TNF molar ration of  10:0. These antibodies are produced by eukaryotic cell lines at greater or equal 100 pg/cell/day.

^*Milestones M1 through M4 shall be deemed met upon delivery of the specified materials to Knoll or an Affiliate thereof and verification of the specified properties.”

One can ignore the eccentric and inconsistent punctuation and spelling in this Schedule. It is agreed by the parties that milestone M4, namely the generation of full antibodies in eukaryotic (i.e. mammalian) cell lines, was part of the program which was to be performed by Knoll. It, rather than CAT, had the relevant expertise in this area. It is not necessary to understand the technical figures in these milestones, it is sufficient to understand that they represent ever greater ability of the fragments to bind to and neutralise TNF.

This brings me to the grant of licences in Article 4/1993. Only some of these provisions need to be referred to as set out below:

“ARTICLE 4

LICENSE GRANT

4.00

CAT grants to Knoll and its Affiliates a world-wide exclusive license under the Patents and Know-How in the Exclusive Filed to manufacture, use, sell, have manufactured, have used and have sold the Product(s) and/or Drug, with the right to sublicense.

4.02

CAT agrees that this Agreement is conditioned on the grant by the Medical Research Council to Knoll and its Affiliates of a sublicense under the Neuberger Patent and the Boss Patent on license terms set out in Schedule III to this Agreement.

4.03

Knoll grants an antigen-based sublicence in the Exclusive Field to CAT under the Neuberger Patent to the extent necessary to permit CAT to conduct its activities under the Research Program, on the terms set out in Schedule IV to this Agreement.”

For completeness I should refer to the definition of the Know-How which CAT licences under Article 4.00/1993. It is to be found in Article 1.14/1993:

“1.14

“Know-how” means all unpatented information, whether or not patentable, relating to materials, methods, processes, techniques and data for the development, manufacture, use or sale of a Product and which any such party is free to transfer or disclose without violating contractual obligations to any third party. Know-how shall include, but not be limited to, regulatory data, market research, marketing strategies and analyses, and other commercial information relating to a Product.”

A number of points arise out of these licensing provisions. First they give effect to the arrangement arrived at between Knoll and CAT to get round the Neuberger and Boss patents. Knoll obtains a sublicence from the MRC for both patents. It needs or may need the Boss patent for the gene expression part of the process (i.e. the part where host cells manufacture full length antibodies). It passes on to CAT a sub licence for the Neuberger patent. This relates to phage display technology. As such it was needed by CAT to enable it to operate the McCafferty phage display invention. Furthermore, it will be appreciated that Articles 4.02/1993 and 4.03/1993 met the requirements imposed on the form of sub-licence by the Celltech/MRC arrangement referred to above. Although it was CAT which secured the relevant rights under Neuberger and Boss, those patents had to be licensed by the MRC to Knoll so that the latter could pass on the sublicence under Neuberger to CAT.

Of more importance are the terms of the main licence. This uses terms which are of a type which is common in this sort of agreement, namely “to manufacture, use, sell, have manufactured, have used and sold the Product(s) and/or Drug”. This entitles the licensee, Knoll, to use the Patents and the Know-How for all and every part of the process, from the very beginning through to the sale of the final commercial product. Thus the Winter 2 library and McCafferty phage display technologies can be used in the early stages of the development program which will lead to the production and sale of such product.

The major dispute between the parties concerns the meaning of Article 5.03/1993 (and the equivalent provision in Article 5.02/1995). It seems to me that it is appropriate to consider this in the context of a number of other provisions in the 1993 Agreement (mirrored in the 1995 Agreement), all of which, to a greater or lesser extent, are interrelated. I set them out below.

“ARTICLE 5

ROYALTIES

5.01

Knoll shall pay to CAT royalty payments based upon the combined Net Sales of Product(s) by Knoll and its Affiliates and sublicensees as follows:

6% of the amount of combined Net Sales of up to one hundred million pounds sterling (£100,000,000) in a given year:

5.5% of the amount of combined Net Sales of over one hundred and fifty (sic) million pounds sterling (£100, 000,000) to one hundred and fifty million (£150,000,000) in a given year;

5% of the amount of combined Net Sales of over one hundred and fifty million pounds sterling (£150,000,000) in a given year.

5.03

Royalties paid to third parties, other than to the Medical Research Council, in order to license rights needed to practice or to have practiced the technology claimed in the Patents, will be borne equally by the parties provided that CAT’s royalty pursuant to 5.01 is not less than 2% (two percent) of Net Sales. The 2% royalty to be paid to the Medical Research Council under the Boss Patent and Neuberger Patent shall be paid directly by Knoll to the Medical Research Council and shall be directly offset against the 4%-6% (sic) royalties due CAT by Knoll.

ARTICLE 6

PAYMENTS

6.00

Knoll shall make payments to CAT with respect to the sale of Product by Knoll, its sublicensees and its Affiliates, on a country-by-country basis, commencing six months after the First Commercial Sale and to continue every six months thereafter. …”

ARTICLE 12

INFRINGEMENT

12.01

If a third party in any country where the Product is being imported, manufactured, used or sold, notifies any of the parties, their Affiliates or licensees that such activity infringes or is alleged to infringe any issued patent either assigned to or licensed to such third party, then:

(a)

the party so notified shall promptly notify the other parties in writing; and

(b)

within thirty (30) days of receipt of notice of the alleged infringement, CAT will notify Knoll that it will either:

(i)

indemnify Knoll and its Affiliates against any and all awards, judgments and settlements which require Knoll or its Affiliates to pay money to such third party and that it will defend the action at its sole expense, or

(ii)

grant Knoll the right to defend, in its sole discretion, the action and allow Knoll to take as a credit, on an ongoing basis during the action, one half the cost of its external legal expenses and any damages which may be awarded, against the royalties due to CAT from Net Sales in the countries in which the claim of infringement is made, up to a maximum credit of 1% of such Net Sales. Such credit may not be taken by Knoll where the action involves a claim or right of a third party based upon activities such as packaging of the Product or improvements to the Product which are beyond the scope of the technology described in the Patents.

12.02

In the event Knoll must pay royalties or license fees to third parties under one or more claims of one or more patents to enable Knoll to utilise or have utilised the inventions of the Patents, half the amount of such payments actually made shall be credited to Knoll’s royalty obligations to CAT for sales in that country where the patents exist. However, Knoll’s royalty obligation to CAT on Net Sales in any given country shall not be reduced below 2% of such Net Sales by operation of the offset of this section 12.02. This offset shall not include royalties or licence fees which are beyond the scope of the technology described in the Patents, for example fees paid to third parties for delivery systems.

ARTICLE 14

WARRANTIES

14.03

CAT gives no warranty, express or implied, that the technology covered by the Patents will work when applied to any particular purpose.

ARTICLE 16

TERM AND TERMINATION

16.04

In the event that Knoll is unable to license on commercially reasonable terms any patent rights necessary to practice or have practiced the inventions disclosed in the Patents, which patent rights would be infringed by the making, using or selling of a Product, Knoll shall have the right, at its election, to terminate the Research Program(s) being conducted hereunder on sixty days written notice.”

For completeness, there are three other provisions in the 1993 Agreement to which I should refer. First is the definition of Net Sales. Insofar as material it is as follows:

“ ‘Net Sales’ mean the amounts invoiced on the sales of Product by Knoll, its sublicensees and Affiliates to independent, unrelated third parties in bona fide arms length transactions, less the following deductions properly documented actually allowed and taken by such third parties and not otherwise recovered or reimbursed by Knoll, its Affiliates or sublicensees…”

This confirms that the royalty provisions of Article 5 above bite on the finished article, in this case HUMIRA. As noted already, that is the basis upon which Knoll has calculated royalties.

Second is Article 21.01/1993 which stipulates that the headings at the beginning of the Articles and Sections of the Agreement are for reference and shall not affect the meanings of the provisions of the Agreement. Third is article 21.07/1993:

“Prior to the execution of this Agreement, the parties have had numerous discussions, conversations and negotiations, and have generated correspondence, writings and other memoranda with respect to the subject matter of this Agreement; notwithstanding which, this Agreement (including its Exhibits) is intended to define the full extent of the parties respective agreements, arrangements and obligations with respect to the subject matter hereof, and each party represents that it is not relying on any such other discussions, conversations, negotiations, correspondence, writings and memoranda in executing and delivering this Agreement or performing its respective obligations hereunder.”

Thus far I have only referred to provisions in the 1993 Agreement. That has now been replaced by the 1995 Agreement and it is under the latter that the royalty payments have been made which have given rise to the current dispute. However, since both parties agree that the 1995 Agreement was intended to have the same effect in all material respects as its predecessor, looking at the provisions of the latter seems to me to be appropriate. Some of the language in the 1993 Agreement was changed to arrive at the language of the 1995 Agreement. Since no one suggests that anything turns on these differences, it is not necessary to set out the terms of all the 1995 equivalents of the Articles set out above. On the other hand, since this dispute centres on the meaning of Article 5.02, read in the light, in particular, of Articles 12.00 and 12.01 of the 1995 Agreement, which are equivalent respectively to Articles 5.03, 12.01 and 12.02 of the 1993 Agreement, I set out below these provisions in the 1995 Agreement:

“5.02

Royalties paid to third parties, other than to MRC under the MRC-Patents, to license rights needed by Knoll, its Affiliates or permitted sublicensees, to practice or to have practiced the technology claimed in the Patents, will be borne equally by CAT and Knoll, provided, however, that the royalty payable to CAT pursuant to Section 5.00 shall not be reduced below two percent (2%) of Net Sales in any year for which a royalty is owed to CAT. The two percent (2%) royalty to be paid to MRC under the MRC-Patents shall be paid directly by Knoll to MRC and shall be directly offset against the 5%-6% royalties due to CAT pursuant to Section 5.00”.

12.00

If a third party in any country where the Product is being imported, manufactured, used , or sold, notifies any of the parties, their Affiliates or licensees that such activity infringes or is alleged to infringe any issued patent wither assigned to or licensed to such third party, then … (b) within thirty (30) days of receipt of notice of the alleged infringement, CAT will notify Knoll that it will … (ii) grant Knoll the right to defend, in its sole discretion, the action and allow Knoll to take as a credit, on an ongoing basis during the action, one half the cost of its external legal expenses and any damages which may be awarded, against the royalties due to CAT from Net Sales in the countries in which the claim of infringement is made, up to a maximum credit of 1% of such Net Sales. Such credit may not be taken by Knoll where the action involves a claim or right of a third party based upon activities such as packaging of the Product or improvements to the Product which are beyond the scope of the technology described in the Patents.

12.01

In the event Knoll or its Collaborators must pay royalties or license fees to third parties under one or more claims of one or more patents to enable Knoll or its Collaborators to utilize or have utilized the inventions of the Patents, half the amount of such payments actually made shall be credited to Knoll’s royalty obligations to CAT for sales in that country where the patents exist. However, Knoll’s royalty obligation to CAT on Net Sales in any given country shall not be reduced below 2% of such Net Sales by operation of the offset of this Section 12.01. This offset shall not include royalties or license fees which are beyond the scope of the technology described in the Patents, for example fees paid to third parties for delivery systems.”

The dispute between the parties centres on the meaning of the words “rights needed to practice or to have practiced the technology claimed in the Patents” in Article 5.03/1993 and Article 5.02/1995. Mr Vos argues that what the parties are taken to have intended by the words they used in these agreements is that the royalty sharing obligation only applies to royalties paid by Knoll/Abbott to third parties for rights which would be infringed by the use of the whole or part of CAT’s antibody gene library and phage display technology. The antibody gene library and phage display technology is what CAT brought to the development programme and in respect of which it had expertise. It is the technology which is the subject of “the Patents”, that is to say Winter 2 and McCafferty. Thus the obligation to share royalties only covers rights of third parties needed to allow the antibody gene library and phage display technology to be used. It is not in dispute that if this construction is correct, none of the deductions from the royalties made by Abbott are allowable.

Lord Grabiner argues for a much wider scope to these words. He says that, were Mr Vos correct, Articles 5.03/1993 and 5.02/1995 would not have used the terminology “the technology claimed in the Patents” but would have used the words “phage-antibody technology”, an expression which had already been used in the Recital to the Agreements, or the rather longer expression “CAT’s antibody gene library and phage display technology” or, perhaps, “CAT’s technology”. Any of those expressions would have made it clear that the dividing line was as now argued for by CAT. The use of the expression “the technology claimed in the Patents” is quite inappropriate to convey the meaning advocated by Mr Vos. But besides attacking Mr Vos’ arguments, he advances positive reasons why the royalty sharing provisions are much wider.

His submission is that the Agreements tell one where the line is to be drawn for the purposes of the sharing provisions. They say that Abbott has no entitlement and CAT has no corresponding obligation to share third party royalties (or legal expenses and damages in Article 12.00/1995) beyond the Product stage and that it is explicit or necessarily implicit that there is an entitlement/obligation down to and including the Product stage. It is important to note that for this purpose Lord Grabiner suggests that “the Product” is the therapeutic antibody expressed from the host cell, not the finished article, HUMIRA. He and Mr Vos accept that the royalty sharing provisions do not cover third party rights throughout the whole of the development up to and including HUMIRA.

Lord Grabiner advances a number of arguments in support of his clients’ contention. First, he says that there can be little room for doubt as to what is meant by the wording used in Articles 5.03/1993, 5.02/1995. The sharing provision is defined by reference to “the technology claimed in the Patents”. The word “claimed” and “Patents” go together and support his construction. As seen above, at the time these words were introduced into the 1993 Agreement, the word “Patents” was defined as meaning both the Winter 2 and McCafferty PCT applications together with those patents when granted, including any patents which derive from them and any divisions, renewals, continuations, continuations-in-part, extensions or reissues. Thus the parties anticipated the future when the applications would mature into granted patents. They did so mature. The granted versions of the McCafferty application include US Patent 5,969,108 (McCafferty). It is admitted by CAT that the method used in the production of the antibodies which comprise HUMIRA and the antibodies themselves fall within claims 19 and 25 of the latter patent. They are so-called reach through claims. Therefore that is the extent of the technology in respect of which CAT is obliged to share the royalty burden. Since the various post-CAT patents, including Cabilly, which are licensed to Abbott cover the production of antibodies which are contained in HUMIRA or steps in such production or the antibodies themselves, the royalties in respect of them must be shared. The same result is achieved if one concentrates first on the word “claimed”. That requires the reader to look at the Winter 2 and McCafferty applications, and the patents granted pursuant to those applications, and examine the scope of all the claims. Once again this leads to the concession made by CAT in relation to claims 19 and 25 of US Patent 5,969,108 (McCafferty). Lord Grabiner argues that Mr Vos’ approach would be a departure from the ordinary meaning of the words used.

Second, he argues that the words “needed by Knoll to practice or have practiced” the technology claimed in the Patent carry with them the meaning that the royalty sharing regime ends with the production or manufacture of an antibody which binds to and neutralises TNF. All of the post-CAT patents on which Abbott/Knoll relies are so needed, in the sense that without them it would not be able to make the antibody. If Abbott can not make the antibody, it can not practice or have practiced the technology in CAT’s patents. The restraint on making the full antibodies has the effect of rendering it pointless to make the small fragments by use of CAT’s library and phage display technology. In that sense, CAT needs the licences under the third party patents to remove the blockage to its work created by the downstream patents. He argues that, were CAT correct in its construction, the words used in the Agreements would have been “needed by CAT to practice or have practiced the technology” claimed in the Patents.

These two arguments are described by Lord Grabiner as his principal ones. However there are others. The third relies on the words “Royalties paid to third parties, other than the Medical Research Council” in 5.03/1993 and their equivalents, “Royalties paid to third parties, other than to MRC under the MRC-Patents” in 5.02/1995. He says that the relevant patents of the MRC include the Boss patent. There is no dispute that this is a patent which relates to a part of the development which is downstream of CAT’s library and phage display technology. This shows that the “third parties” being contemplated by these provisions in the Agreements include owners of rights subsequent to the phage display step. It is bound to cover Cabilly and its patent which is essentially to the same invention as in Boss. This is put particularly clearly in the Re-Amended Defence:

“8(2) … The MRC-Patents (including Boss) were examples where royalties would be paid to third parties in order to license rights needed by Knoll to practice or have practiced the technology claimed in the Patents. They were treated differently from other third party rights only in the sense that a full 2% deduction was to be made from the royalties payable to CAT.

12(4) … This construction accords with the purpose and intent of the 1995 Agreement and is consistent with the fact that the MRC-Patents (including Boss) are expressly stated to be examples where royalties would be paid to third parties in order to license rights needed to practice or have practiced the technology claimed in the Patents.”

Fourth, he says that express exclusions from the sharing obligations set out in 12.00/1993 and 12.01/1993 and the 1995 Agreement equivalents, namely “activities such as packaging of the Product or improvements to the Product” and “for example fees paid to third parties for delivery systems” both indicate that the sharing provisions do not go all the way to the final retail product, namely HUMIRA in its blister-packed final dosage form, including hypodermic syringe, but, on the other hand, indicate that it is only the very final stages, after the manufacture of the pharmaceutical grade full antibody from host cells which are excluded from the reckoning. If CAT’s construction were right, the examples of exclusions would explicitly cover stages in the development closer to the phage stage, for example the stage of creating the full length antibody, so as to throw light on the dividing line between those things which are the subject of the sharing obligation and those which are not.

Before considering these arguments, I think there are some simple and, so I understand it, non-contentious matters which may throw some light on the true construction of these provisions. First, I understand both parties to agree that, in principle, the parties could have chosen to make CAT’s obligation to share royalties as wide or as narrow as they wished. It would have been possible to agree to make CAT share royalties paid to third parties in respect of rights necessary to allow its library and phage display technology to be used but no further. This, of course, is what CAT says the Agreements do. On the other hand it would have been possible, and as commercially rational, to have decided that, since CAT was to receive royalties based on the receipts made by sale of HUMIRA in its final dosage form, it should share the burden of all third party royalties incurred in reaching that stage. In other words it should share the burden of all and any third party licences taken by Knoll. It is to be noticed that this is not what Lord Grabiner argues is the effect of these provisions. As an alternative, the parties could have chosen a boundary for the royalty sharing obligations between these two extremes. That is what Lord Grabiner is arguing for.

Following on from that, it is worth noting that there is a compelling reason why Abbott does not argue that Articles 5.03/193 and 5.02/195 provisions oblige CAT to share the burden for all third party licences. It is not in dispute, in fact it is one of Lord Grabiner’s submissions, that the royalty sharing obligation in 5.03/1993 has identical scope to those in 12.02/1993 and that the same applies to 5.02/1995 and 12.01/1995. 12.02/1993 and 12.01/1995 state expressly, but by way of example, that royalties on delivery systems are not to be offset, that is they are not to be part of the sharing arrangement. Different examples of what are excluded are to be found at the end of 12.01/1993 and 12.00/1995. Although neither 5.03/1993 and 5/02/1995 expressly refer to downstream stages being excluded from the sharing obligation, to be consistent with the provisions in Article 12, they must be construed that way. So, the royalty sharing agreement does not cover the whole of development, production and sale. As I say, this is agreed by the parties. Lord Grabiner’s arguments are designed to show that the boundary of liability is immediately after the stage of production of full length pharmaceutical quality antibodies. I understand his argument to be that such a boundary can be discerned in the language of 5.03/1993 and 5.02/1995. It is not dependent upon the terminology of the express exclusions mentioned in Article 12. In other words, he is not expressly or implicitly importing into the former any of the wording from the latter. The exclusions mentioned by way of example in Article 12 are merely illustrative of what is inherent in Articles 5.03/1993 and 5.02/1995

Both sides also agree that the drafting of the Agreements is not always internally consistent. In particular it is agreed that the following expressions are to be interpreted as having the same meanings:

I entertain no serious doubt that Mr Vos’ construction of the royalty sharing provisions is the correct one. It appears to me that it is the only construction which is consistent with all the other provisions of the Agreements and, in the factual matrix within which the Agreements are set, makes commercial sense. Indeed, the more one considers Lord Grabiner’s arguments, the more difficult his client’s position appears to be. I will consider each of Lord Grabiner’s arguments in turn.

It seems to me that the arguments which are based on the precise dictionary or patent law meaning of the words “claimed” and “Patents” in the relevant Articles do not accord with the modern approach to construction. I have set out above passages from Lord Hoffmann’s speech in the Investors’ Compensation Scheme case. Lord Grabiner also alluded to another House of Lords authority, namely Kirin-Amgen Inc v. Hoechst Marion Roussel Limited [2004] UKHL 46. However, the following extract from the speech of Lord Hoffmann in that case emphasises that a pedantic, dictionary-led approach to construction is not correct:

“32.

Construction, whether of a patent or any other document, is of course not directly concerned with what the author meant to say. There is no window into the mind of the patentee or the author of any other document. Construction is objective in the sense that it is concerned with what a reasonable person to whom the utterance was addressed would have understood the author to be using the words to mean. Notice, however, that it is not, as is sometimes said, "the meaning of the words the author used", but rather what the notional addressee would have understood the author to mean by using those words. The meaning of words is a matter of convention, governed by rules, which can be found in dictionaries and grammars. What the author would have been understood to mean by using those words is not simply a matter of rules. It is highly sensitive to the context of and background to the particular utterance. It depends not only upon the words the author has chosen but also upon the identity of the audience he is taken to have been addressing and the knowledge and assumptions which one attributes to that audience. I have discussed these questions at some length in Mannai Investment Co Ltd v Eagle Star Life Assurance Co Ltd [1997] AC 749 and Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896.”

Furthermore if one follows the route suggested by Lord Grabiner, one arrives at a construction of the relevant Articles which is inconsistent with the rest of the Agreements and makes no commercial sense. As Mr Vos points out, the relevant words in Article 5.03/1993 are agreed to mean the same as the equivalent words in Article 14.03. Thus the latter Article would have to mean that CAT is giving no warranty that any technology used in Abbott’s plant up to and including the antibody production phase works. Article 14.03 cannot mean that. It clearly is only concerned with whether or not CAT is giving warranties in respect of its own library and phage display technology. If the words mean that in Article 14.03/1993, they must mean the same in 5.03/1993 (and their equivalents in the 1995 Agreement).

This is not the only problem created by this construction. As I have said already, two obvious ways of defining the royalty sharing burden would be either to make it apply only to CAT’s technology or to make it apply to the whole of the development and production process up to and including the marketing of the finished product. Both of those have the advantage of being certain and easy to understand. Abbott is arguing for a dividing line somewhere in between these two, but the construction advanced would make that line moveable or worse. The 1993 Agreement was entered into when there were no granted patents. As Lord Grabiner emphasises, the parties anticipated that there would be granted patents in the future and he relies on the claims in one of them to indicate the scope of the royalty-sharing obligation. But that means that the extent of that obligation will vary according to the wording of the claims in the different patents granted and from country to country. Indeed, even on the wording of US Patent 5,969,108 it appears to me that the finished product, HUMIRA, is an infringement. If so, the royalty sharing obligation, at least in that country, would cover the whole development and production process – something which Lord Grabiner is anxious to avoid. At its extreme, this would have the effect of giving an incentive to CAT to limit the wording of the patent claims thereby limiting the royalty-sharing obligation. I should make it clear that this would have no significant effect on the scope of protection achieved by CAT. There is a well known illustration of this in patent folklore. The inventor of a steam whistle who claims, in the first claim in his patent, the whistle and, in a later claim, a battleship to which the whistle is attached, obtains no greater rights by doing so. He can not have the battleship delivered up or destroyed even if his invention of the whistle is a good one. In any event, if the scope of the obligation on CAT is measured by reference to the terms of the claims in any or all of its patents, it would mean that the borderline would vary from time to time and for country to country. Presumably, were Lord Grabiner right, CAT could remove its obligation to pay royalties for any post-phage stages by simply surrendering any claims which purport to cover such stages.

This leads on to another difficulty with this approach to construction. Neither of the Agreements contains any warranty that either of the patent applications cover a patentable invention, or that valid patents will be secured. They do not oblige CAT to pursue the patent applications through to grant or to seek wide claims nor do they provide that the failure to secure patent protection is a ground for termination. Abbott’s obligation to pay royalties for the use of the technology which CAT had introduced and in relation to which it had all the expertise is simply not dependent upon the existence or scope of CAT’s patent protection. If, for example, CAT failed to secure patent protect or surrendered any protection it had, there would be no claims to carry the royalty sharing provisions. If it transpired that another patentee had enforceable patent rights in relation to the CAT library or phage display technology (cf the overlap between the Boss and Cabilly patents), Abbott would be obliged to pay licence fees to that patentee and to CAT but, because CAT would have no patents and, therefore, no claims, Abbott could not recoup any part of the royalties from CAT. In my view, this would make commercial nonsense of the transaction.

Perhaps anticipating this type of problem, Lord Grabiner softened this argument in his closing written submissions. He now says:

“Abbott’s case is that when construing Article 5.02, the question is what is claimed in the Patents. But, as one might expect, the technology described in the Patents does not differ significantly from what is claimed in the Patents.”

However, moving the emphasis away from “claimed” to “described” does not make the argument any stronger. It may make it worse. First, it destroys any semblance that construction is based on the precise dictionary meaning of the words in the relevant Articles. Second, it makes the dividing line between cases in which CAT must and those were it need not share royalties at least as obscure and variable. Since the definition of Patents in the Agreements covers patents, including improvements patents, which may be applied for in the future, once again the scope of CAT’s obligation and Abbott’s entitlement will vary depending the wording of the descriptive part (i.e. the specifications) of different patents in different countries, many of which may not have been in contemplation at the time the Agreement was entered into. What the impact would be if none of the patents were granted or if CAT were to surrender them is unclear. Furthermore, if one inquires what is the “the technology described in the Patents”, one is left with a field which is even wider than if one concentrates on the word “claimed”. Nearly every patent starts by describing prior art technology to which it lays no claim. Presumably this is technology “described in the Patents”. Further it is normal to describe reagents and equipment used before, during and after the part of the process which the patentee has invented. Is the royalty sharing obligation intended to relate to these items as well?

It seems to me that these arguments would render the meaning of the royalty sharing provisions obscure and fluctuating. They would make no business sense.

I turn to Lord Grabiner’s second principle arguments namely that which concentrates on the meaning of the words “needed by Knoll”. This can be broken down into a number of parts. First, he says that the words in the Article carry with them the meaning that the royalty sharing regime ends with the production or manufacture of an antibody which binds to and neutralises TNF. As mentioned already, this is not the retail product, HUMIRA. He says that the former, not the latter, is the Products as defined by the agreement. Further, he argues that, were CAT’s construction correct, the words “needed by Knoll” should have been “needed by CAT”.

There are a number of reasons why I do not accept any of these arguments. First, the Agreements suggest that HUMIRA falls within the definition of Product, not outside it as suggested by Lord Grabiner. I have set that definition out above. I have also explained why I do not accept that the term “Product” goes no further than the isolated antibodies. It is not necessary to repeat those reasons here. Furthermore, if Lord Grabiner were correct, one would expect to see a reference to the Product in the royalty sharing Articles. But neither that term nor any reference to the final antibody is to be found there. It seems to me that this argument is little more than assertion.

The other argument is one which Lord Grabiner puts forward with more force. However it appears to me that it is untenable and, if anything, undermines his construction. What Lord Grabiner says is that, if CAT’s construction were correct, the royalty sharing Articles would all refer to “needed by CAT” rather than “needed by Knoll”. It follows that CAT must be wrong. When fully expanded to display the steps in the logic, what is being said is the following. First, the words “needed by” identify which of the parties requires the licence. It is common ground that the development programme leading to HUMIRA involves two separate areas of responsibility. CAT is responsible for the part of the process up to and including phage display. It carries that part out in Cambridge, England. Knoll/Abbott is responsible for the post-phage stages up to and including the manufacture of HUMIRA. That is carried out in the United States. If CAT’s argument that the royalty sharing obligation only extends as far as the phage display step, the Agreements would have said that the licences were “needed by CAT”, because that was the part of the development for which it had sole responsibility.

Once expanded like this, it will be seen that the argument does not support Abbott’s construction and would make the royalty sharing provisions irrational. As far as the latter is concerned, at the heart of Lord Grabiner’s submission is the argument that, because the development is in two parts with each part being the responsibility of one party only, CAT’s obligation to share royalties must apply to the post-phage stages. Those are the stages where Knoll/Abbott, rather than CAT, may “need” third party licences. However, if that is what the Article means, it would also follow that the royalty sharing provisions do not apply to the library and phage display stages of the development because, on Lord Grabiner’s argument, those are stages where only CAT would “need” third party licences. The result would be to turn these Articles on their heads and to require CAT to share third party royalty obligations only in the areas of the technology for which it is not responsible. No one has suggested that this is what these Articles mean. They would make no commercial sense if they did.

Furthermore, even if Lord Grabiner were correct, it would not help him to show, as he must, that the royalty sharing obligations finish with the creation of the therapeutic antibody, not HUMIRA. Third party rights may exist up to and including the final delivery system in which HUMIRA is sold. If Lord Grabiner were correct, third party rights acquired anywhere in these post-phage display stages would be “needed by Knoll” and therefore within the royalty sharing obligation. As everyone agrees, that cannot be what the Articles mean because the Agreements say expressly, but only by way of example, that parts of the development, such as delivery systems, are not covered by the royalty sharing provisions.

In my view, the words “needed by Knoll” throw no light on the extent of the royalty sharing provisions. In particular, I do not accept that they are inappropriate to cover only the library and phage display parts of the development. Although CAT is to do the work in this area, the whole development, including these stages, are being done for and on behalf of Knoll/Abbott. Commercially, Knoll is as responsible for these stages as is CAT. Were third parties to sue in respect of the use of their patented technology in, say, the phage display part of the programme, Knoll would be as liable as CAT. Indeed, the fact that Knoll needs the licences in these stages is put beyond argument by the fact that both of the Agreements includes licences to Knoll under the Winter 2 and McCafferty patents – something which would make no sense if Knoll does not need licences for these stages.

I can therefore turn to Lord Grabiner’s other two arguments. First I should consider his reliance on the references to the MRC licences in Articles 5.03/1993 and 5/02. It will be recalled that this is based on the fact that the MRC licence covers Boss. Therefore, it is argued, owners of rights in technology downstream of phage display, such as Boss, are relevant third parties for the purpose of these provisions. Agreements under their rights must be covered by the obligation to share royalties.

I agree with Mr Vos that the natural meaning of these Articles is that the MRC licences are not examples of the general royalty sharing obligation but are to be treated sui generis. In addition, Lord Grabiner’s arguments cause problems. First, they do not assist him in demonstrating where the border line between sharing and not sharing is. True enough, if he were right, the sort of rights covered by the Boss patents would fall inside the net. But there is nothing to indicate how much further the net goes. Second, this construction would put a great strain on the wording of Articles 12.02/1993 and 12.01/1995. It will be recalled that it is common ground that the latter Articles must be of identical scope to Articles 5.03/1993 and 5.02/1995 respectively. However there is no reference in the former two Articles to the MRC Patents and the licences under them. It follows that if, as Lord Grabiner asserts, these patents are within the area covered by the royalty sharing obligation, under Articles 12.02/1993 and 12.01/1995, half of the royalties would be borne by CAT whereas under Articles 5.03/1993 and 5.02/1995, all of the royalties would be so borne. Of course this problem does not arise if, as Mr Vos argues, the royalties under the MRC licence do not fall to be treated as part of the royalty sharing regime but are being dealt with sui generis.

Lord Grabiner’s other argument is that the express exclusions from the sharing obligations set out in 12.00/1993 and 12.01/1993 and the 1995 Agreement equivalents, namely “activities such as packaging of the Product or improvements to the Product” and “for example fees paid to third parties for delivery systems” are to be expected to be close to, and therefore illustrate, the border between technology for which CAT has to shoulder some of the royalty burden and technology for which it does not. Were CAT’s construction correct, one would have expected to see the examples of excluded matter much closer to phage display. For example, expression from the host cells would have been mentioned as an excluded technology.

This argument is dependent upon the examples in these Articles being a long way down stream from phage display and therefore indicating, indirectly, that the royalty sharing borderline is likewise a long way downstream of this point. It is for that reason that Lord Grabiner pays particular attention to the reference to packaging in Article 12.00/1993 and delivery systems in Article 12.01/1993. However the terms of Article 12.00/1993, if anything, point away from Abbott’s construction. The examples of excluded matter includes “improvements” to the Product. Lord Grabiner suggests that all such improvements must be after the formation of the full length antibody. In my view there is nothing to support such a suggestion. Indeed, it seems to me to be contrary to such indications as there are in the Agreements. It will be recalled that “Product” is defined as meaning any pharmaceutical composition containing one or more of the “Antibodies” and that Antibodies are defined as meaning “antibody molecules or fragments or derivatives thereof, identified, produced or derived under the Research Program binding to The Antigen (emphasis added). I can see no reason why improvements in the fragments from which Products are eventually derived or improvements to the expression stage (for example incorporating different levels of glycosylation or substituted amino acids in the constant regions) should not be covered by the wording of these Articles. In other words, I do not accept that these exclusions indicate that the border line is where Lord Grabiner suggests. If it points anywhere, it supports CAT’s construction. However, it seems to me that this is really too small a point to throw light on the true meaning of the royalty sharing provisions.

I have set out above at some length the reasons why I do not accept Abbott’s construction of the Agreement. By contrast, Mr Vos’ construction not only is consistent with the wording used but makes commercial sense. CAT only brought its library and phage display technology to the programme. It did not profess to have any expertise outside these fields. In particular it was given no remit to assist or even offer views as to how expression and later stages in the production of HUMIRA were to be conducted. The latter were all to be carried out by Knoll in its premises in the United States. The Agreements did not contain any obligation on Knoll to notify CAT of the technology it would use or contemplate using in the post-phage stages. It was not under an obligation to notify CAT of the third party rights it was considering taking licences under. It was not under any obligation to discuss the royalty rates it was contemplating accepting. In the circumstances it made obvious sense that CAT’s obligation to offer at least partial protection to Knoll against the impact of third party licence demands should be limited to the areas of the program for which it, CAT, was responsible. Whereas the dividing line suggested by Mr Vos makes commercial sense, I am not persuaded that Lord Grabiner’s does. As explained above, there are sensible reasons why it might be appropriate for CAT to share the royalty burden in relation to its technology only or in relation to the whole program from the final results of which it would obtain its royalty. However it is difficult to see any rational basis upon which it could have been decided that CAT should share the royalty burden on part only of those parts of the technology which were to be Knoll’s exclusive responsibility. None was suggested by Lord Grabiner.

In my view this is also consistent with the wording used in the Agreements. I have noted above that the parties agree, correctly, that the expressions “the technology claimed in the Patents” in Articles 5.03/1993 and 5.02/1995, “the inventions of the Patents” in Articles 12.02/1993 and 12.01/1995, “the technology described in the Patents” in Articles 12.02/1993, 12.01(b)(ii)/1993, 12.01/1995 and12.00(b)(ii)/1995, “the technology covered by the Patents” in Articles 14.03/1993 and 14.02/ 1995 and (5) “the inventions disclosed in the Patents” in Article 16.04/1993 all mean the same thing. For this reason it is even less appropriate here than in other cases to approach the words “technology claimed in the Patents” in a pedantic and acontextual way. The expressions used in all these Articles point to a common idea, namely the novel technology which is described in Winter 2 and McCafferty and forms the inventive heart of their claims. This is, as Mr Vos says, CAT’s library and phage display technology.

In the light of this finding, it is not strictly necessary to consider Mr Vos’ backup arguments of rectification and estoppel. However it is possible that this case may go further and, in the light of that possibility, I think it would be appropriate for me to express a view, particularly since these grounds involve, at least in some respects, the assessment of the witnesses who gave oral evidence.

Mr Vos runs two groups of argument. First he says that, were Lord Grabiner’s construction of the royalty sharing provisions correct, the Agreements would not accord with the true intentions of both CAT and Knoll. He says that this would be a case of common mistake and that the licences should be rectified so as to express clearly that CAT’s obligations only cover third party licences in the library and phage display parts of the programme. Fairly late in the day, after it had received Knoll’s witness statements in August of this year, it amended its pleadings to add another ground, namely unilateral mistake. Mr Vos explains that, if some of the defendant’s evidence is believed, particularly that of a Mr Salfeld and Mr Shaw, it would appear that Knoll misled CAT into thinking that Knoll shared CAT’s belief as to the meaning of the royalty sharing Articles whereas, in truth, Knoll believed that it meant that CAT had a much more extensive obligation. It is said that, on this premise, Knoll behaved unconscionably in not disabusing CAT of its misapprehension. This would be a case of unilateral mistake which would justify rectification. Alternatively, on essentially the same grounds, Knoll would be estopped from advancing its construction of the Agreements.

The unilateral mistake and estoppel arguments can be disposed of speedily. They are both dependent upon Knoll having behaved unconscionably during the 1993 and 1995 negotiations and of having had, and knowing it had, a different view of the breadth of the royalty sharing provisions to that held by CAT. Having read the witness statements and listened to the witnesses, there is no basis upon which either of those arguments could succeed. On the contrary, both sides agree that the discussions between them were amicable and open. They were conducted without either side trying to hide what it thought from the other. The parties went out of their ways to ensure that they understood each other’s position. I do not think there is any evidence on which CAT can rely that Knoll tried to pull the wool over its eyes. On the contrary, as will be seen from what follows, each side fully understood what the other wanted and what it thought the provisions in the various drafts passing between them were meant to cover. In particular in the area which was crucial to CAT, namely its royalty entitlement and the extent of any obligation to share third party royalties with Knoll, there was complete clarity as to what was wanted by CAT and what was agreed. This, of course, is of central significance to the common mistake plea, but it takes the wind out of the unilateral mistake and estoppel arguments.

In his Opening Skeleton, Mr Vos said that the relevant principles of law relating to rectification of contracts on the ground of common mistake are well settled. They have recently been have been summarised recently by Lawrence Collins J in T&N Ltd (in administration) & Ors v. Royal & Sun Alliance plc & Ors [2003] EWHC 1016 (Ch) at paragraphs 133 to 137:

“133.

There was no dispute as to the applicable principles for rectification … In the case of common or mutual mistake the party seeking rectification must show that the concluded instrument does not represent the common intention of the parties, and must show some outward expression of accord or evidence of a continuing common intention outwardly manifested: Frederick E Rose Ltd v. William H Pim & Co Ltd [1953] 2 QB 450, 461-462 (CA); Joscelyne v. Nissen [1970] 2 QB 86, 97 (CA).

134.

There has to be outward accord for otherwise - “there could be no certainty at all in business transactions if a party who had entered into a firm contract could afterwards turn round and claim to have it rectified on the ground that the parties intended something different”: (Denning LJ in Frederick E Rose (London) Ltd v. Wm H Pim & Co Ltd [1953] 2 QB 450 at 462).

135.

The party seeking rectification does not have to meet more than the civil standard of balance of probabilities, but convincing proof is required to counteract the cogent evidence of the parties’ intention displayed by the instrument: Thomas Bates and Sons v. Wyndham’s Ltd [1981] 1 WLR 505, 521 (CA); Grand Metropolitan plc v. William Hill Group [1997] 1 BCLC 390 at 394.

136.

Rectification may be available if the document contains the very wording that the parties intended it to contain, but it has in law or as a matter of true construction an effect or meaning different from that which was intended: Re Butlin’s Settlement [1976] Ch 251, 260; Grand Metropolitan plc v. William Hill Group [1997] 1 BCLC 390 at 394. Contrast Harlow Development Corp v. Kingsgate (Clothing Productions) Ltd (1973) 226 Est Gaz 1960 (CA).

137.

It is not necessary that the parties should at the material time have formulated the words which it is sought to insert by rectification. It is sufficient that the parties had the necessary common continuing intention as to the substance of that which would be achieved by the rectification sought: Crane v. Hegeman-Harris Co Inc [1971] 1 WLR 1390 at 1399; Grand Metropolitan v. William Hill Group, at 394-5”.

There did not appear to be too great a difference between the parties, at least in opening. Lord Grabiner’s position was that the Court has power to rectify an agreement if the parties had a continuing common intention down to and including the execution of the written agreement which, by common mistake, was not expressed in the agreement. He accepted that it was unnecessary to show that there was in place an enforceable contract immediately prior to the execution of the written document, but there must have been an “outward expression of accord”. He said that it was trite law, first, that the subjective musings of one, or other, or indeed both parties which did not “cross the line” in the course of the negotiations will not provide any or any adequate basis for a claim to rectify a contract. Secondly, the subjective opinion of a party as to the purpose or intent of an agreement was also not good enough. By these submissions, I understood him to mean that the subjective intentions of the parties, by themselves, was not enough to found a case of common mistake. In addition, there had to be the outward expression to which Lawrence Collins J refers.

Since there did not appear to be any dispute between the parties on these issues, it was no surprise that both put in evidence as to their intentions, and their understanding of the other side’s intentions, during the negotiations leading up to the two Agreements. Reference was also made to various documents and discussions passing between the parties in which their respective intentions were set out. Much of the evidence put in by CAT, mainly in the witness statement of its Dr Chiswell, was not challenged in cross-examination. However all of Knoll’s witnesses, including in particular Dr Salfeld, Dr Müller-Neumann and Mr Shaw, were subjected to careful and detailed cross-examination by Mr Vos.

In his speech after the cross-examination was finished, Lord Grabiner explained that he did not rely on the evidence of what Dr Salfeld or Mr Shaw said they understood to be the parties’ respective and mutual intentions at the relevant time. This evidence was, he said, irrelevant and inadmissible as was the evidence on this subject given on behalf of CAT, primarily by Dr Chiswell and Mr Nixon. He puts it as follows in his written closing submissions:

“Abbott’s case is that the only material which is legally admissible for the purpose of deciding whether or not an agreement or common understanding was reached is that which crossed the line. The internal ruminations of the parties or either of them are inadmissible and irrelevant: what matters is whether and, if so, how they have “outwardly manifested” their alleged agreement or common intention. Significant time was taken up at the trial by counsel for CAT reading out to witnesses extracts from a very large number of documents which, for the most part, never crossed the line. Those documents and the opinions of witnesses as to what those documents meant or what their respective thought processes were at the time are simply inadmissible and therefore irrelevant to the legal inquiry which has to be conducted by the Court in the case of common mistake rectification.”

Lord Grabiner advances this argument with some force. He says that any other view is heresy and that he is incontrovertibly correct. Indeed he goes as far as to say that he is prepared to go to the stake on this point. One cannot but be impressed by this level of selfless dedication to his client’s cause. I am not sure whether he expects the rest of the Abbott legal team to join him on the pyre, but, a single raised eyebrow from Mr Thorley suggests that he, at least, is not going to volunteer. I hope Lord Grabiner can be persuaded to think again, otherwise it would be the end of what has the signs of being a promising career. In spite of Lord Grabiner’s threat of self immolation, or perhaps encouraged by it, Mr Vos relies on numerous authorities and references in Snell and Spry to show that Lord Grabiner’s view of the law is wrong. He also argues that, even if he were right, there are here so many documents which “cross the line” and support CAT’s case, that it succeeds anyway.

As Mr Vos says, and in this he is supported by Snell’s Equity (13 Ed 2000), rectification is an equitable remedy designed to circumvent a potential defect in the common law, namely the enforcement of a written instrument which does not accord with the true intention of the parties. By deploying this remedy, courts of equity rewrite the instrument so that it accords with the parties’ true intention. Amongst other things, the party seeking the relief must prove a continuing common intention up to the time the instrument is executed. As Snell puts it:

“Any difference between the instrument and the terms previously agreed may be due not to mistake but to a decision by the parties to depart from those terms. A claim for rectification must therefore establish that the parties intended the instrument merely to carry out those terms and not to vary them. If one of the parties denies that the instrument was contrary to his intention this accordingly has considerable weight; but like other evidence in cases of rectification, this evidence must be directed to the intention existing when the instrument was executed, or earlier, and not at some later time.”

It seems to me that central to this is the need to show what the parties intended at the relevant time. Because the law, as it stands at present, also requires an outward manifestation of the common intent, the intentions of the parties separately and together are not sufficient to secure relief. But, as Snell shows, the intention of the parties and evidence of it, is clearly relevant. This is supported by much authority. I will only refer to a few. First, it is worth remembering what was said by Lord Hoffmann in the Investors’ Compensation Scheme case (supra):

“The law excludes from the admissible background the previous negotiations of the parties and their declarations of subjective intent. They are admissible only in an action for rectification. The law makes this distinction for reasons of practical policy and, in this respect only, legal interpretation differs from the way we would interpret utterances in ordinary life. The boundaries of this exception are in some respects unclear. But this is not the occasion on which to explore them.” (emphasis added)

Central to this passage is the distinction drawn between the common law approach to construction and the equitable doctrine of rectification. In the former one does not look at the subjective intent of the parties, in the latter one does. Similarly in BCCI v. Ali [2002] 1 A.C. 251, Lord Nicholls said:

“This judgment in the High Court of Australia in Grant’s case contained also the observation that the surrounding circumstances to be taken into account include the state of knowledge of the respective parties concerning the existence, character and extent of the liability in question and “the actual intention of the releasor”. For many years the accepted wisdom has been that evidence of the actual intention of the parties is not admissible on the interpretation of a written agreement, although such evidence is admissible for other purposes, for example, on a claim for rectification. In Investors Compensation Scheme Ltd v. West Bromwich Building Society [1998] 1 WLR 896, 913, my noble and learned friend Lord Hoffmann pointed out that the exclusion from evidence of the previous negotiations of the parties and their declarations of subjective intent is for reasons of practical policy. He added that the boundaries of this exception are in some respects unclear.” (paragraph 31, emphasis added)

The same point is made in Joscelyne v. Nissen [1970] 2 QB 86, a case relied on by Lord Grabiner in support of his proposition. There the Court of Appeal cited with approval its previous decision, Lovell and Christmas Ltd v Wall 104 LT 85 and, in particular, the following passage from the judgment of Buckley LJ:

“In ordering rectification the court does not rectify contracts, but what it rectifies is the erroneous expression of contracts in documents. For rectification it is not enough to set about to find out what one or even both of the parties to the contract intended. What you have got to find out is what intention was communicated by one side to the other, and with what common intention and common agreement they made their bargain.” (emphasis added)

Once again, this emphasises the need to show the parties intentions outside the terms of the written instruments although those intentions, without more, are not enough to justify rectification. The Court of Appeal also referred with approval to the judgment of Simonds J in Crane v Hegeman-Harris Co. Inc. [939] 1 All ER 662. Among the passages cited was the following:

“The judge held, and I respectfully concur with his reasoning and his conclusion, that it is sufficient to find a common continuing intention in regard to a particular provision or aspect of the agreement. If one finds that, in regard to a particular point, the parties were in agreement up to the moment when they executed their formal instrument, and the formal instrument does not conform with that common agreement, then this court has jurisdiction to rectify, although it may be that there was, until the formal instrument was executed, no concluded and binding contract between the parties. That is what the judge decided, and, as I say, with his reasoning I wholly concur, and I can add nothing to his authority in the matter, except that I would say that, if it were not so, it would be a strange thing, for the result would be that two parties binding themselves by a mistake to which each had equally contributed, by an instrument which did not express their real intention, would yet be bound by it. That is a state of affairs which I hold is not the law, and, until a higher court tells me it is the law, I shall continue to exercise the jurisdiction which Clauson J., as I think rightly, held might be entertained by this court.”

The whole thrust of that passage is to distinguish the objectively assessed meaning of the words used in an instrument from the “real intention” of the parties. Rectification is relief intended in defined circumstances to give effect to the “real intention” when it differs from objective construction. It seems to me that it is impossible to determine what the real intent of the parties is without admitting evidence of what they intended.

Many other authorities were relied on by Mr Vos. I do not need to refer to more of them. I reject Lord Grabiner's argument.

Before considering the evidence adduced on this issue, there are one or two matters which will help in understanding the history of the communications between the parties. First, I have already mentioned that Knoll was part of the BASF group of companies until it was purchased by Abbott in 2000/2001. For most purposes, it makes little difference whether reference is made to Knoll or Abbott. In 1992 and 1993, the environment for conducting development work in the genetic engineering field was more hostile in Germany than it was in America. As a result, BASF set up a new American company, BASF Bioresearch Corporation (“BBC”), to carry out research and development work in this field in the United States. It was that company which, to a large extent, conducted front line negotiations with CAT. Although it was, by BASF standards, a small and young company, it represented the negotiating arm of both Knoll and BASF. No agreement could be entered into without the agreement of the boards of the latter two companies.

During most of the negotiations, CAT’s negotiating team consisted of two individuals, Dr Chiswell and Mr Nixon. Dr Chiswell was the Chief Executive Officer of CAT. He was the founder of CAT in conjunction with Sir Gregory Winter (one of the inventor of the Winter 2 patent). He had the prime responsibility for negotiating the Agreements. He resigned from CAT in March 2002. Mr Nixon is a qualified barrister. He was formerly the Company Secretary and the Vice President of Operations and Legal Affairs of CAT but in 1991 he formed and became a director of The Warwick Partnership Limited. It was in that capacity that he provided in-house legal services to companies, including CAT. In effect, CAT had a part time or locum legal department in the form of Mr Nixon. Mr Nixon is not a scientist. Such limited knowledge as he has of the technology involved in this dispute he has learned only as a result of his performance of legal services on behalf of CAT.

The negotiating team on the Knoll side was much larger, as one might expect from a long established and experienced pharmaceutical group like BASF. There were many people who were involved, directly or indirectly, with the negotiations. Only some of them gave evidence. The more important personalities are as follows.

Dr Joachim Kahlstorff was the in-house lawyer at Knoll from 1984 to the end of 2003. He was responsible for drafting the 1993 Agreement and, as a member of Knoll's team, involved in negotiating the 1993 Agreement. He was, with Professor Schlick, a signatory to both Agreements.

Dr Jochen Salfeld is an important witness. He was a Senior Scientist at BBC and Research Project Leader on the TNF Project from January 1993 onwards following Dr Klaus Schollmeier’s return to Germany. He was a member of the contract negotiating team, co-ordinating input from the BBC team. He was enthusiastically in favour of Knoll entering into an agreement with CAT and took over from Dr Schollmeier as leader of the negotiating team. He is now the Vice President, Biologics Discovery at Abbott Bioresearch Centre.

Another major participant was Dr Markus Müller-Neumann. He was a member of Knoll’s Licensing Department from Spring 1992 to Summer 1993. He had previously worked in Knoll’s Research Planning/Project Management Department from 1989 to Spring 1992, primarily on TNF and MAK-195 projects. He is a scientist by training.

Professor Eric Schlick was the first President of BBC until late 1990. He was head of R&D at Knoll from January 1991 and member of the Knoll Board. He left BASF in June 2000. He was a signatory, with Dr Kahlstorff, to the 1993 and 1995 Agreements. He was not called as a witness.

Dr Robert (Bob) Kamen was President of BBC from June 1991 onwards. He was previously R&D Director at Genetics Institute Inc. and Research Scientist at the Imperial Cancer Research Fund. He is the author or recipient of a number of significant documents in the court bundle. He is now retired. He was not called to give evidence. CAT asserts that Abbott’s solicitors, in effect, prevented them from interviewing him.

Dr Thomas Dörper has been a Patent Attorney at BASF in Ludwigshafen, Germany since October 1990 working in the Biotechnology and Pharmaceutical section of the Patent and Licensing Department. He was previously a biochemist at BASF for 5 years and worked on Knoll’s first anti-TNF murine antibody project. He reported to a Dr Bieberbach. He was also the co-author of a patent opinion dated 3 June 1993 to which reference will be made below. He was involved at the time of both the 1993 and 1995 Agreements.

Mr Robert Shaw was a US qualified attorney. He was Vice President, Intellectual Property and in-house attorney for BASF from 1989 until July 1997. He was involved in drafting the 1995 Agreement, taking his instructions from BBC. He was also a patent attorney and one of his claimed areas of expertise was in relation to licensing.

Dr David Berstein was the in-house Patent Attorney at BBC from April 1990 to September 1993. He was based in Worcester, Massachusetts, USA. He commented on drafts of the 1993 Agreement and on patent issues. He gave a witness statement in these proceedings in effect saying that he could remember nothing. In the circumstances Mr Vos did not require his attendance for cross examination.

There were other individuals on the Knoll side. As and where necessary, I will explain their position in the hierarchy.

There is one other matter which needs to be borne in mind. In its Re-Amended Defence, Abbott refers to the various third party patents, the licences to operate which it says it is entitled to share with CAT. These patents are set out in two Schedules, A and B to that pleading. The first contains a number of patents under which Abbott says it has incurred royalty obligations which are sufficiently large that, once shared with CAT, they reduce CAT’s royalty entitlement to the minimum, 2%. Those in the second Schedule are other patents which Abbott claims to be entitled to treat in the same way. Since CAT has already been pushed down to the minimum, it has not been necessary to take them into account. However Abbott says that they are available to be taken into account, if, for example, one of the Schedule A patents expires so that royalties are no longer due under it. The patents in these two Schedules are referred to in the documents and in the submissions as Cabilly, Aggarwal, Wallach, Axel, Crowe, Keen, Page and Rathjen.

In determining what the intention of the parties was and what they were agreeing to, the primary source is evidence of witnesses to the negotiations and the contemporaneous documents. The negotiations leading to the 1993 Agreement took place 11 years ago. Those relating to the 1995 Agreement took place 9 years ago. It is inevitable that the passage of time will have blurred the recollection of some of the participants. For that reason the contemporaneous documents, drafted without an eye to litigation and at a time when, according to all the witnesses, the parties were getting on well and trusted each other, are particularly important. After all, as someone once said, it is more reliable to have a short pencil than a long memory.

As mentioned above, right through to the late 1990’s the correspondence and the meetings between the parties were open and neither side tried to pull the wool over the other’s eyes or slip something past the other. As far as the negotiators themselves are concerned, I have no doubt that all of them took their tasks seriously. But there was at least one significant difference between the negotiating teams. Those on the Knoll side were experienced operators in the pharmaceutical field. They were, in effect, BASF personnel with all the considerable financial muscle of that company behind them. Furthermore they were negotiating as to whether BASF were to pay 2% or about 5% to CAT. The difference between the two was 3% of their total hoped for income from sales of the products of the development program. For CAT the story was rather different. It was a new company with very few employees. It did not sell pharmaceuticals. It traded in its research expertise. For it, the discussions concerning royalties were to determine whether its total income from this licence was to be 2% of Knoll’s turnover in HUMIRA or two and a half times that figure. This negotiation was crucially important to CAT and, as the leader of the small negotiating team, crucially important to Dr Chiswell. It is not surprising that he was able to recollect the negotiations clearly. Dr Chiswell was an utterly fair and reliable witness. I have no hesitation in accepting his evidence. Of course his task was not arduous since, in many crucial areas, his evidence was supported by the contemporaneous documents.

My conclusion was that Mr Nixon was also a reliable witnesses on the issues he could recall. However he was not a scientist and did not appear to have the day to day involvement that Dr Chiswell had. In those circumstances it is not surprising that in some areas he could not recall events or discussions and had to try to reconstruct events as best he could from the documents.

A number of Knoll witnesses were called. Some of them were subjected to lengthy, precise and telling cross-examination by Mr Vos. At times I suggested to Mr Vos that this could have been dealt with more rapidly but, in the event, his thoroughness left little room for doubt as to the reliability or otherwise of the Knoll witnesses and its version of the negotiations.

Dr Salfeld played a major role in the negotiations with CAT, particularly in relation to the 1993 Agreement. He prepared three witness statements for this action. In paragraphs 3 and 4 of his first witness statement in which he said:

“3.

From early 1993 until 2000, almost 80% of my working day was devoted to various aspects of the collaboration with CAT including the contractual arrangements for, and the research and development of, both D2E7 (Humira®) and J695 (a monoclonal antibody to IL-12). Even today I still spend probably one third of my time on these CAT-related matters. I have some good recollections of events surrounding the drafting and negotiating of the 1993 Agreement, although only from January 1993 when I became a member of the negotiating team. I was less involved in the drafting and negotiating of the 1995 Agreement, but remember certain issues that arose in that context.

4.

I have read documents from the files at BBC (including my own files concerning the contracts with CAT and the patent issues), the files at Knoll, the files provided by BASF Corporation and CAT’s List of Documents. Not all of the documents look familiar. Furthermore, I do not remember the exact events and details recorded in some of the documents. Where relevant, I have been asked to refer, in this statement, to documents that will help explain events. I have tried to indicate where the documents themselves and/or their contents do not form part of my personal recollections.”

The overall impact of this and the rest of the evidence, as written, was that he had a fairly good recollection of many stages in the negotiations, particularly in 1993, even if he did not remember all the documents and the little details. He expressed strong disagreement with CAT’s version of events. His written evidence also gave the clear impression, which I believe is likely to accord with the facts, that he had read all or most of the relevant documents recently for the purpose of preparing his own witness statements and answering specific points made by CAT in its pleadings and evidence. He was subjected to a particularly lengthy and detailed cross examination. As it proceeded, and particularly when Mr Vos put to him documents which appeared to be inconsistent with his version of events, he professed to have an ever smaller recollection of dates, places, people and events and virtually no recollection of any of the documents. By the second half of his cross-examination, he professed to be unable to recall virtually anything. He suggested on a number of occasions that he was reading documents for the first time in years, yet they were obviously documents which he would have had to consider for the purpose of preparing his witness statements and in analysing CAT’s evidence. It is not necessary for me to determine the cause of this progressive amnesia. It does not matter. What became clear at quite an early stage was that Dr Salfeld’s evidence of what was or was not discussed during the negotiations and what was or was not Knoll’s understanding of the agreements or their appreciation of CAT’s understanding was completely unreliable. Many of his answers appeared to be inconsistent with contemporaneous documents. On any issue where there was a difference of recollection between Dr Chiswell and Dr Salfeld, I accept the evidence of Dr Chiswell.

The other witness on the Abbott side who stood out as particularly unreliable was Mr Shaw. His evidence, like that of other witnesses, was to the effect that the purpose of the 1995 Agreement was to repeat the provisions of the 1993 Agreement but with minor drafting changes to clean it up and with alterations to cater for two new targets. Other than in the latter respect, the version of the 1995 Agreement as signed was intended and believed to have exactly the same effect as its predecessor. He was only involved with the drafting of the 1995 Agreement. As such, his evidence was not central to most of the issues before me. However Mr Vos argues that the behaviour of Knoll in relation to the drafting of the 1995 Agreement confirms that Knoll knew all along that the royalty sharing provisions only extended to third party rights which covered CAT’s library and phage display technology and did not cover down-stream stages such as expression of the antibodies. As such, it did not provide for royalty sharing for patents such as Cabilly. Further it confirms that Knoll knew that that was CAT’s understanding. It was in relation to these points that Mr Shaw gave two witness statements and was cross-examined by video link, ill health encountered since he made his witness statements having prevented him from attending the trial in person.

I have to bear in mind Mr Shaw’s health concerns. This may have made him less open and relaxed. I am sorry to say that, this caveat notwithstanding, he came over as a determined and unflinching witness who was not convincing. He struck me as a man who had a story to put across and who was not prepared to let contemporaneous documents get in his way. Where such documents did not accord with his version of events he simply asserted that they did. On a number of occasions when the apparent conflict between his evidence and the documents cried out for clear explanation, he simply made assertions and brushed aside the inconsistencies. I came to the conclusion that his evidence, particularly where it conflicted with the evidence of others or documents, could not be relied upon. Unfortunately, it is difficult to put this unreliability down to lapses of memory. The clear impression given to me was that Mr Shaw was giving inaccurate evidence and he knew it.

Although it is out of sequence, there is one issue covered by Mr Shaw’s evidence which it is convenient to deal with here. The true story of Mr Shaw’s involvement in the negotiating process was that, when he read the 1993 Agreement he immediately recognised the limited scope of the royalty sharing provisions in Article 5.03. As a result, he produced a revised draft of the Agreement for submission to CAT. In that, Article 5.03 was redrafted by Mr Shaw as follows:

“5.03

Royalties paid to third parties, other than to MRC pursuant to the MRC-Knoll-Agreement to license rights needed by Knoll, its Affiliates or permitted sublincesees, to make, use or sell Antibodies, Product(s) or Drug, will be borne equally by CAT and Knoll, provide, however, that the royalty payable to CAT pursuant to Section 5.01 shall not be reduced below one percent (1%) of Net Sales in a any year for which a royalty is owed to CAT. The two percent (2%) royalty to be paid to MRC under the MRC-Patents shall be paid directly by Knoll to MRC and shall be directly offset against the royalties due to CAT pursuant to Section 5.01” (emphasis added)

In fact this was never forwarded to CAT. It was stopped internally in Knoll. Mr Shaw’s oral evidence was that the change from “rights needed to practice or to have practiced the technology claimed in the Patents” in 5.03/1993 to “rights needed by Knoll, its Affiliates or permitted sublincesees, to make, use or sell Antibodies, Product(s) or Drug” in his draft was merely clarification. He said that it never occurred to him that the wording in the 1993 Agreement meant anything different to his proposed wording. I have no doubt that in relation to this issue, Mr Shaw’s evidence is not reliable. On the contrary, if this had merely been a matter of clarification, it would have been put to CAT and probably been adopted in the 1995 Agreement, as many other insubstantial drafting changes were. In my view it is overwhelmingly likely that Mr Shaw realised the limited scope of Article 5.03/1993 and wanted it expanded so that Cabilly, which by this time was a significant concern to Knoll, would fall within the obligation to share royalties. It was stopped inside Knoll because Knoll appreciated at all times that CAT was only prepared to accept a sharing obligation in relation to its own technology, that this is what Knoll had agreed to and that CAT would not accept a change.

Returning to the subject of the witnesses. There was one more Abbott witness who deserves to be mentioned at this stage. He is Dr Müller-Neumann. He appeared to me to be uncomfortable during his cross examination. My impression was that he wanted to be as accurate as he could, but on a few occasions he felt no choice but to support a version of events with which he was not happy. The result was that, save in respect of one or two issues, he managed to steer clear of controversy.

It is important to bear in mind that this part of my judgment proceeds on the assumption that my finding as to construction is wrong and that Abbott is right in its assertion that Articles 5.03/1993, 12.02/1993, 5.02/1995 and 12.01/1995 all impose on CAT an obligation to share third party royalties in respect of technology extending far downstream to the phage display stage. If that is so, CAT says that it is clear that the parties always intended, understood and agreed that the sharing was only to be in respect of CAT’s library and phage display technology and the wording of those Articles must be rectified accordingly. CAT’s evidence that this was its understanding, as set out fully and explicitly by Dr Chiswell and Mr Nixon in their witness statements, was not challenged under cross-examination. They were also clear that, to their knowledge, this was Knoll’s understanding as well. There is one passage in the witness statement of Sir Gregory Winter (who Lord Grabiner did not wish to cross-examine) which adequately illustrates CAT’s position on this issue.

“34.

The issue of the royalty offset in relation to the TNFa project was discussed at Board level within CAT. This was an important project and these were clearly key commercial issues. I do not remember seeing the final legal agreement, but the commercial principle that we should be responsible for sharing the pain for patents that might cover our technology was fully discussed and accepted. These sorts of issues were often discussed in pre-meetings of the Board, and the Board meetings themselves were reserved for formal decisions.”

There was no challenge to the assertion of the importance of this issue to CAT. In my view it is inconceivable that it was not discussed openly with Knoll. Furthermore it is not a complicated issue. All that was needed was agreement on the location of the borderline of liability.

However, in the witness statements served on its behalf, Abbott denied that the understanding of Knoll, BASF or any of the individuals who conducted the negotiations on their behalf was the same. It says it believed at all times that the royalty sharing was to cover the downstream technology as well. It was in relation to this that I was taken through a very large number of documents, both internal and passing between the parties, and there was the extensive cross examination to which I have referred already. It does not seem to me appropriate or necessary to go into this in the detail which Mr Vos, understandably, did. I will concentrate on some of the major steps in the negotiations before, and the behaviour of the parties before and after, the Agreements were executed. But I can summarise my findings now. In my view the case put forward by CAT is overwhelmingly supported by the documents and evidence. It is quite clear that CAT never intended to accept royalty sharing in relation to post-phage technology. It is clear that Knoll knew that and the Agreements were entered into with the intention that it should accord with that mutual understanding and intention. Consistent with this, both before and after the Agreements were executed both parties acted consistently with that understanding until Abbott came to make the first royalty payment.

The story can be taken up in mid-1992. By this time Dr Chiswell had made a presentation to Knoll. CAT made it clear that it wanted a 6% royalty on the net sale price of the products made as a result of the use of its technology. By this time CAT had also discussed the possibility of securing a sublicence to Neuberger and Boss through its close association with the MRC. On 28 July, Dr Chiswell sent Mr Berstein the Pat Confid document in which CAT set out its understanding of the whole of the patent field which might impinge on the production of antibodies by a process involving phage display. It splits the technology into three sections. The first is called “Pre-phage (Library construction). The second is described as “Phage Display (Selection)”. The third is “Post-phage (Manufacturing)”. The relevant patents in each section are described. The first refers to Winter 2. The second refers, of course, to McCafferty but it also refers to other patents relating to this stage, including Ladner, Affymax and Neuberger. The third stage also refers to a number of patents including, in particular, Boss and Cabilly. As far as the latter patent is concerned, it points out that there is an interference between that and Boss and notes:

“It is likely that Cabilly will win some aspects of the interference, at least for chimaerics, because of the ability to demonstrate invention from notebooks”

So, from the very outset, Knoll was aware that there was at least the possibility that, in the United States, Cabilly would prevail in some respects over Boss. Cabilly continued to be a concern for Knoll over the next 11 years. Not once in all that time, until it made the first royalty payment, did it suggest that CAT had agreed to share the burden of any royalties due under it.

The separate roles CAT and Knoll would play in the development of an anti-TNF antibody were discussed between the parties. There was no dispute that CAT’s role would be in relation to library and phage display technology only. There was a meeting between the parties on 7 October 1992 at which the possibility of CAT being “flexible” over royalties was discussed. This resulted in Dr Chiswell sending a letter to Dr Müller-Neumann on 14 October setting out CAT’s revised position. It is important to the issues I have to decide. For the first time, reference is made to the possibility of royalty sharing. It is CAT’s proposal and reads as follows:

“The risks of additional royalties being paid to third parties because of licences needed to practice CAT’s technology to be shared. Royalties will be reduced by 50% of third party royalties down to a minimum of 5%.” (Emphasis added)

Dr Chiswell’s evidence on this was clear. It meant that what was being offered was royalty sharing for third party rights but only in relation to CAT’s technology. I do not see how this letter could be read in any other way. In fact Dr Chiswell’s evidence was that at this stage what was meant to be offered was sharing in relation to the McCafferty technology of phage display only. At this time there was no intention to offer to share any third party royalty burden in other parts of the putative process, even for the library stage of Winter 2.

There is no dispute that this was considered carefully by Knoll. The result was a counter offer dated 30 October1992. It was signed by both Dr Kahlstorff and Dr Müller-Neumann. The latter told me that a double signature like this meant that it was an official document of the company. It contained what must have been a carefully considered response. The letter sets out in three columns the various proposals made by CAT and Knoll’s response to them. In particular it contains the following wording:

It will be seen that CAT’s proposal contained three elements. First, it would share the royalties paid to 3^rd parties for rights to practice its own technology. Second, the sharing would be at the rate of 50%. Third the minimum royalty would be 5%. The Knoll response appears to me to be clear. First, it accepted that CAT should have responsibility for licences to third parties to practice CAT’s technology. The wording of this is lifted straight from CAT’s proposal with two amendments; the addition of the words “to have practiced” – which no one suggested added anything of substance – and a change from sharing to 100% responsibility for such third party royalties with no minimum. It also makes clear what was inherent in CAT’s proposal, namely that all other royalties due to third parties should be borne by Knoll. The other matter added by Knoll was a reference to the MRC royalties (i.e. for Neuberger and Boss). But this is clearly treated as quite separate from the sharing regime under discussion.

Lord Grabiner suggests that the omission from the second bullet in the third column of the word “develop”, which is to be found in the first paragraph of that column, necessarily implies that it is to be included in the first bullet. Thus the counter proposal was that CAT should have responsibility for the royalties incurred in respect of third party rights which covered its technology and development technology.

In my view there is no justification for any such implication. On the contrary, the natural meaning of this part of the letter is that Knoll is accepting CAT’s suggestion that it should bear responsibility for third party royalties in respect of its own technology but no further. The difference is only as to whether that responsibility should be as to 50% or as to 100%. Lord Grabiner’s suggestion was firmly rejected by both Dr Chiswell and Mr Nixon under cross-examination. It was not credibly supported by any of the Knoll witnesses.

Indeed, it seems to me that Lord Grabiner’s argument would have produced obviously unacceptable results. It will be remembered that the counter offer was that CAT was to shoulder all of the relevant third party royalties and that there would be no minimum royalty for CAT. If Lord Grabiner were right, Knoll were suggesting that CAT were to accept 100% responsibility for royalties in respect of areas of the technology used by Knoll for which it, CAT, had no responsibility and over the choice of which it had no input. In almost any foreseeable scenario, CAT would have ended up with no royalties at all for its ground breaking technology, even if there were no competing patents in the library and phage display fields. I do not believe that Knoll ever intended any such thing.

In addition, there is a conspicuous lack of rationale to Abbott’s construction. If the purpose of this wording was to make CAT bear the third party royalty burden in respect of “developments”, why would it make sense to exclude marketing? CAT had no more knowledge of or input into one than the other. Furthermore, if development was to be included, that would seem to go much further than Lord Grabiner is now arguing for. Indeed the only thing excluded would be marketing. This would mean that things like delivery systems could come within the sharing obligation. That is something which has not been suggested on behalf of Abbott. Finally, as I have said already, the evidence is that the relationship between CAT and Abbott was open, fair and co-operative. If Knoll had really intended to include greater responsibilities for CAT, there is no doubt that it would have said so clearly, not by a subtle, almost invisible, play on words like this.

If there were any doubt on this matter, it is resolved by Dr Chiswell’s letter in response dated 5 November which is entirely consistent with the evidence he gave orally during the trial. This letter deserves analysis. It is almost entirely concerned with the counterproposal which Knoll had put forward concerning the burden of third party royalties. The importance of its content both to CAT and Knoll is apparent. First it deals generally with Knoll’s suggestion that CAT should bear 100% of the third party royalty burden in relation to its, CAT’s, technology. Dr Chiswell wrote;

“Our difficulty with your suggestions concerning third party royalties is, however, a major one. Your proposal could lead to a situation where CAT had provided the know-how for the products and our patents had issued but, nevertheless, CAT would receive no royalties.

…However, biotech patents being what they are, there can be no certainty about any position until very late in the day. We therefore have to establish a reasonable view of the patent risk associated with this technology. CAT is certainly prepared to take some of the patent risk for its technology whilst avoiding situations where our partners can effectively negotiate away our royalties in the event that even a small piece of the technology requires another licence.” (Emphasis added)

There can be no serious dispute as to what this meant. CAT was prepared to shoulder part of the burden for its own technology. Not the whole burden. There is no suggestion whatsoever that it was prepared to shoulder any part of the burden for third party licences to technology other than its own, nor that it understood that Knoll was suggesting any such thing. Needless to say, Knoll never said that Dr Chiswell had misunderstood its position. There is little doubt it would have done so had Dr Chiswell’s understanding been incorrect.

The letter goes on then to consider various patents. It splits them into three groups; (i) Boss and Neuberger, (ii) Winter 2 and (iii) McCafferty. For each Dr Chiswell suggests a notional royalty of 2%. In relation to (i) and (ii) he points out that the patents are comparatively old so that the risk of finding some unknown problem is low. As for McCafferty, he says that this is a newer application. For that reason, the risk of unknown problems is higher. Based on this assessment he then proceeds as follows:

“We propose dealing with these groups separately such that royalties due from the low risk patents are not subject to any reduction. That is 2% from Neuberger/Boss plus 2% from Winter 2. CAT would then shoulder the whole risk arising from third party licences needed to practice McCafferty et al.”

Once again the meaning of this is clear. There would be no assumption of the burden of third party royalties in respect of any other technology – even Neuberger/Boss and Winter 2. The only burden which would be accepted related to the use of the McCafferty technology, that is to say phage display. In fact the letter goes on to modify the latter suggestion. Instead of sticking with bearing the whole of the burden for McCafferty, Dr Chiswell wrote:

“We propose the approach whereby the CAT royalty rate due to McCafferty et al is reduced by 50% of any third party royalties due to licencees needed to practice this part of the technology. Thus if third party royalties are 2%, McCafferty royalties would be reduced to 1% but Knoll’s overall royalty rate would be increased from 6 to 7%; if third party royalties were 4% McCafferty royalties would be reduced to zero but Knoll’s overall rate would have been increased to 8%. This approach allows CAT to feel that it is in Knoll’s interests to negotiate hard and in good faith

To summarise, if these proposals are agreed the CAT royalty rate would be 6% but that this would be reduced by 50% of third party royalties needed to practice McCafferty et al until the CAT royalty rate of 4% was reached.” (emphasis added)

I do not see how there could have been any misunderstanding as to what Dr Chiswell was saying. The royalty sharing would extend only to third party rights in relation to the phage display technology of McCafferty. Dr Müller-Neumann accepted as much under cross-examination. That Dr Müller-Neumann understood correctly what Dr Chiswell was suggesting was confirmed during a conversation between the two on 19 November. This is consistent with Dr Müller-Neumann’s internal note of that conversation and was confirmed under cross-examination.

The next significant document is a Knoll internal note from Dr Salfeld to Dr Dörper. It deals with what it terms “post-CAT antibody modifications”. It is an analysis of the routes being contemplated by Knoll for the production of full antibodies after CAT produced the correct scFv fragments. It includes an analysis of third party patents which might prove a hurdle. Interestingly, this contains a number of references to the Axel patent which is now said to be one of the ones which Knoll is entitled to set off against the royalties to CAT. Although these patents were discussed, they were never mentioned to CAT. Indeed, this is but one of numerous documents in the disclosure which demonstrate that Knoll considered the impact and possible royalties due under third party patents relating to the post-CAT part of the process leading to Humira. However, according to the documents and the credible evidence, not once over the ten years until these proceedings commenced did Knoll suggest to CAT that any royalties incurred by Knoll in respect of them would fall to be shared. Indeed, in nearly all cases Knoll did not even mention to CAT that it was thinking of taking a licence under such rights. As shall be seen below, even when Cabilly was mentioned to CAT, it was not to suggest that there was an obligation to share the royalties due, but rather to enlist CAT’s help in seeking to persuade the MRC to reduce its 2% royalty for the Boss/Neuberger rights in order that Knoll could better afford to pay the Cabilly royalties. This is only consistent with one version of events, CAT’s. It is telling evidence that Knoll had exactly the same understanding as to what CAT wanted and it had agreed to as CAT did.

On 16 November, Knoll sent another letter to CAT. Again it was signed by both Dr Kahlstorff and Dr Müller-Neumann. It contains the following:

“The only concern we still have is handling the risk and splitting of third party royalty payments. Nevertheless we will strongly consider this issue as addressed by you. We have transferred your information concerning third party licenses to our patent department which will work out their opinion as quickly as possible”

Once again, this emphasises that the question of splitting the royalty burden was of high priority to both parties. In the light of this, Dr Salfeld’s inability to recall what was discussed on this topic is surprising. Dr Chiswell’s clarity of recollection is not. In view of its importance, it would be very surprising if the parties did not understand precisely what each other was saying. I have come to the conclusion that they did understand.

The first draft of the proposed licence was prepared by Knoll. Internal copies of the latter marked in manuscript by various of the Knoll witnesses show that Article 5.03 was drafted to take into account Dr Chiswell’s letter of 5 November, referred to above. Of particular interest is the copy of the draft marked up by Dr Dörper. Against the proposed Article 5.03 he has written “why not 50% for Winter 2”. Once again, there is no doubt what this meant. Dr Dörper considered that if CAT’s liability to share third party royalties was to be limited to those relating to its own phage display technology only, it should be expanded to cover the Winter 2 library technology since this also came from CAT. This was accepted by Dr Müller-Neumann under cross-examination.

Various other changes were made to the draft licence, including stylistic changes to Article 5.03. None of them changed the sense of the latter. The provisions of Article 12 were also made consistent with it. Furthermore Article 5.03 was amended by the insertion of a reference to the Boss/Neuberger licence to the MRC. It is clear that this was added without in any way changing the sense or extent of the royalty sharing provisions.

Although there are additional documents which Mr Vos relies on, it is sufficient to move on to 8 February 1993 on which day there were two meetings between Knoll and CAT. One related to patents, the other related to contract matters. The latest draft of the licence was being considered. Mr Nixon had gone through the draft and had marked up any clause which needed consideration. He produced a list for discussion which was faxed through to Knoll. Against Article 5.03 he wrote:

“this Article should make it clear that it refers to the technology covered by the Patents only, not any technology covered by other patents referring to the target.”

Mr Nixon’s note was used as a checklist during the meetings. It is apparent that CAT wanted to make sure that there was no doubt that its royalty sharing obligations extended no further than the “Patents” – that is to say the library and phage display technology. There can be little doubt that this subject was discussed. Some of the Knoll participants’ notes of the meetings record nothing against this item. Dr Kahlsdorf’s simply says “OK”. Dr Salfeld’s note says “fine”.

Knoll put in evidence that, had there been a substantial change in these provisions at this time, that would have been recorded on their minutes. I agree. But the point was that there was not a substantial change. What Mr Nixon wanted clarified was what was already by this time the mutual intention of the parties. There was little to say. Everyone agreed that CAT’s liability was limited to the library and phage display technology. This topic must have been considered and dealt with very quickly.

In his witness statement Dr Chiswell gave evidence he recalled discussing Article 5.03 with Knoll representatives on one occasion and that he infers, from his review of the documents, that it must have been at the contract meeting on 8 February. He said:

“the outcome was that we agreed with Knoll that it was clear that the scope of the royalty sharing provision in this article was limited to CAT’s technology only, not any technology covered by other patents referring to the target or to antibody formatting or to manufacturing methods; in other words, the article was clear as drafted and did not need to be amended.”

Understandably, this evidence was challenged by Lord Grabiner in cross examination. He points out that Dr Chiswell could not remember the location of this discussion nor the people who were present. This evidence is, he suggests, unreliable.

Lord Grabiner’s criticisms have to be assessed in the context of what was going on at this time. The evidence is that the contract meeting was protracted. There were some issues which were the subject of major discussion. The royalty sharing provisions was not one of them. All the rest of the evidence shows that this must have been disposed of in a very short time. It is not in the least surprising that Dr Chiswell can remember that there was agreement without being able to recall the words used or the participants. I have read and re-read the relevant passages of Dr Chiswell’s cross-examination. My view now is what it was during the trial. Dr Chiswell’s evidence was compelling and credible. I accept that such a discussion did take place, that agreement was expressed as Dr Chiswell says and that it almost certainly took place during the contract meeting on 8 February 1993.

Lord Grabiner accepts that the parties agreed something in relation to Article 5.03/1993 at the 8 February meeting, hence the “OK” in Dr Kahlsdorff’s notes. But he says that all that was agreed was the form of works, not what they meant. I do not accept that. Mr Nixon’s note and evidence show that CAT wanted the limits of its responsibility to be clear. In my view, on the evidence, it is virtually certain that this was achieved and agreed between the parties at the meeting..

I should mention one other contemporaneous document from this time which throws light on the intentions of the party. When Mr Nixon received his copy of the first draft of the agreement he paid particular attention to Article 12.02. There is no dispute that the parties wanted this to be consistent with Article 5.03. In the margin he wrote:

“only if infringement occurs [because] using McCafferty/Winter 2”

In his witness statement, Mr Nixon explains this as follows:

“This illustrates my understanding that a royalty offset would only arise where there would be an infringement of a patent owned by a third party as a result of using the technology of the McCafferty and Winter 2 patents.”

This was not challenged under cross examination. It goes to reinforce the point that CAT was always clear as to the limit it was insisting on with respect to sharing third party royalty obligations. Article 12.02 was one of the articles which was down on Mr Nixon’s checklist used as the agenda for the contract meeting on 8 February. Again this reinforces the view that this would have been a subject discussed and agreed at that meeting.

On 9 February, Dr Müller-Neumann produced a draft minute of the patent meeting of the previous day. It refers to the Affymax and Protein Engineering (Ladner) patents. Then it goes on to deal with what are described as “Other patents needed ‘post CAT’ (Cabilly, Boss, etc)”. He wrote:

“Both routes BBC intends to follow (complete antibody or fragment) will be covered by Boss, i.e. we will have access to these rights via CAT. There is, however, an interference between Boss and Cabilly in the USA. If Cabilly succeeds, there is a risk of facing broad claims in the USA. In Europe Cabilly does not touch us, since we do not use chimeric antibodies.

This issue is not a specific CAT problem, but it concerns many other companies as well due to the fact that the basic biotechnology know-how is involved.”

Dr Müller-Neumann was cross examined on this, particularly in relation to the second paragraph. It was suggested that this amounts to a written acknowledgement that Cabilly was not something with which CAT was concerned, that is to say that it was outwith the royalty sharing provisions. Dr Müller-Neumann disputed this. He said that he thought it was referring to the fact that Cabilly was a problem whether or not CAT technology was used. It is not possible now to resolve this dispute. On the other hand the first paragraph is just as important. It is but one of a number of occasions on which the potential importance of the Cabilly patent was commented upon within Knoll. As noted already, not once did Knoll suggest to CAT that this patent came within the royalty sharing provisions until the commencement of this dispute.

This point is made particularly clearly in the next couple of documents to which reference should be made. First there is a manuscript note by Dr Dörper of a meeting he had with Dr Salfeld and Dr Berstein on 11 February. It contains a section headed “post-CAT-Technology”. It refers to a number of patents and assesses the royalties which would be payable by Knoll were licences to be taken. The patents include Cabilly and Axel (both patents now said to fall within the royalty sharing arrangement and listed in the Schedules to Abbott’s pleadings). Against none of them is there a suggestion that royalties will be shared. Once again, notwithstanding this consideration of Axel within Knoll, it was never mentioned to CAT as a possible candidate for sharing until after this dispute commenced.

The second document is from Dr Berstein to Dr Dörper and Dr Müller-Neumann. It is a comment on the last mentioned manuscript note. It contains a number of paragraphs of significance. First it says:

“The patent environment was reviewed with regard to the intended CAT-BBC project. Potentially relevant patent applications are held by Affymax, Protein Engineering Corp, Celltech (Boss) and Genentech (Cabilly). All patent issues involving Affymax would be resolved if a proposed business relationship between CAT and Affymax is consummated. Risks from Protein Engineering Corp are difficult to quantify at present because the scope of patent claims likely to be granted to PEC is difficult to predict, especially in Europe. Both Affymax and PEC patent risks would be shared by CAT and Knoll under the proposed agreement. While Knoll would have a license under Celltech’s Boss patent, it is possible that in the US (only) broader relevant claims may issue to Cabilly (Genentech). The likelihood that such claims will be granted to Genentech is difficult to assess at the moment since relevant documents are kept confidential by the patent office during the pendency of a Boss-Cabilly interference.” (Emphasis added)

There is no dispute that the Affymax and Protein Engineering (i.e. Ladner) patents covered the library and phage display part of the technology. They therefore fell within the scope of the third party rights which CAT was prepared to share. That is what this paragraph says in terms. But, by contrast, although Cabilly is noted as a potential risk, there is no suggestion of royalty sharing in relation to it. The document then goes on to discuss the Affymax and Protein Engineering (Ladner) patents and, after doing so, has a section with the rubric “Patents relating to ‘post CAT’ aspects of the project”. In the latter, again, reference is made to Cabilly and the potential royalty consequences to Knoll. What it does not suggest is that any of such royalties would be shared with CAT. It appears clear from this that royalty sharing was only being contemplated for the first two patents, not Cabilly. Under cross-examination Dr Müller-Neumann appeared to accept that that was probably his understanding at the time. In my view there can be little doubt that that is what this document assumed. At this time Knoll had no belief that royalties on post-CAT technology was to be shared. An official document containing just this distinction between Affymax and Ladner on the one hand and post-CAT Cabilly on the other, was then signed by Dr Müller-Neumann. It reflects Knoll’s understanding of the limit of CAT’s sharing obligation.

Mr Vos argues that consensus as to royalty sharing arrived at during the meetings between the parties on 8 February remained unchanged up to and including the execution of the 1993 Agreement. I am not aware that this was disputed by Abbott. In any event, Mr Vos is correct on this. In fact it seems to me that the fundamental principle that CAT would not bear any royalty sharing obligations save in relation to the use of its own technology had been made clear at the very beginning of the negotiations and remained right through to the end. Both parties understood and agreed that. I would only add that the documents suggest that it is probably because so many people had an input into the drafting of the Agreement that the different wording was used for the same concept in Articles 5.03/1993, 12.02/1993, 12.02/1993, 12.01(b)(ii)/1993, 14.03/1993 and 16.04/1993. They were never intended to have different scope.

Although there was no subsequent change in the parties’ respective intentions or understandings, there is one other document from this period to which I think I should refer. Before the 1993 Agreement could be entered into, approval for that course had to be obtained from the Boards of Knoll and BASF. For that reason, an analysis of the pros and cons of entering into the Agreement had to be put before the Boards so that they could determine what to do. The analysis prepared in BBC included a patent review dated 3 June. Like the documents referred to above, it covered both Affymax and Ladner, in respect of which it noted that the “risks [of royalties] would be shared by CAT and Knoll under the proposed agreement”. It also covered what are described as “Patents relating to ‘post CAT’ aspects of the project”. This refers, amongst other things, to Cabilly and Axel. Again, there is no suggestion of any royalty sharing in relation to these patents. If, as Abbott now asserts, it had always been the intention, at least of Knoll, that royalties due on such patents would be borne equally by CAT, it is inconceivable that this would not have been spelt out to the Boards, particularly since the sharing in relation to the other two patents was spelt out. The fact that it was not mentioned is yet another indication that CAT’s case on this point is correct and Abbott’s is not.

I should add that when this was put do Dr Müller-Neumann under cross examination, he suggested that it was “self evident” that there was to be sharing in relation to post-CAT rights. I think Mr Vos is right to criticize this evidence. It does not accord with the obvious meaning of the documents or the history leading up to the Board’s approval. Dr Müller-Neumann went on to give the following evidence:

“A. Well, I mean, reading this document now, and interpreting what I said is difficult, but I think for me, and I assume for the team as well, it was clear that we had to share, let us say, all the burden until the antibody is there, and then of course share the benefit, hopefully 10 years later.

Q. I suggest to you that that was never your intention, Dr Muller-Neumann, and never the intention of Knoll?

A. Well, sorry, sir, this is what you propose, but I would like to reject this because there is no proof of this and there is no proof of that. This is just what you say. I can just tell you my impression we had and what was our approach, to be open to all the people as good as we can, to incorporate everything we could, to share all the burden and to share all the benefits and to have at the end of the project a product.

Q. The proof was, Dr Muller-Neumann, that on 8th February 1993, you had put forward a draft which simply extended the royalty sharing from McCafferty to CAT's patents, meaning McCafferty and Winter II, and you had never suggested to CAT, and you have never suggested to CAT throughout the negotiations in 1993, that downstream patents would be shared with them?

A. I mean, without -- if I may say, Mr Vos, I am pretty sure about one thing, that when we went to CAT in the very first meeting in October, and certainly in the course of the negotiations, we must have talked about these basic genetic engineering patents, because we very freshly had this experience with TNF and Genentech. It was so fresh, not only in my mind, but I am sure also in the team's mind, that we certainly have raised this right in the beginning. So if you say we did never discuss this, this is what you state, I would say we must have discussed this, because it was evident, genetic engineering is always needed and it was also needed in this project. So if it is not specifically mentioned, maybe it was clear for everybody that it was so, and maybe we have a document from CAT rejecting that, I do not know, but for me it was clear, and I am sure for the team as well. Somewhere I have seen when we went yesterday through these documents there was a quite strange looking document which was hardly readable -- I do not know where it is in the stack -- where it seems that the team as well is agreeing there is more than these CAT patents. We have this broad approach from the very start of the idea to the final antibody, and that is the project, and whatever we need, we can clarify as good as we can, but there are some questions open. The basic patents were always an issue, I am very sure on that. If I remember one thing, only one thing, it is that -- what I said this morning -- "ceterum censeo" -- we always said: let us not forget about all this genetic engineering stuff we need for every project. We can see this in Bieberbach's correspondence and Kahlstorff, and so on.

Q. Dr Muller-Neumann, you have no recollection of ever suggesting to CAT that basic genetic engineering downstream patents would be shared with them, do you, no recollection, as you sit here today?

A. I have no specific recollection on this, but I am very sure that we talked about it because it was apparent that we needed these rights, so we must have talked about this internally repeatedly, and I do not see any reason why we should not have mentioned this in the discussion with CAT. This would not make sense because ...

Q. Dr Muller-Neumann, I am asking you very specific questions. I will just ask it once again: you have no recollection of discussing sharing downstream royalties with CAT at all, do you, actual recollection?

A. What do you mean downstream?

Q. Sharing basic genetic engineering patents like Cabilly and Axel and others with CAT -- you have no actual recollection of saying to CAT, "We are going to share Cabilly, we are going to share other basic genetic engineering patents"?

A. I do not have a specific recollection, we talked about it at a specific date whatsoever, but I am pretty sure that we mentioned this issue during the course of the three-quarters of the year.

Q. I suggest to you, Dr Muller-Neumann, that that was never suggested and that the negotiations proceeded as I have put to you over the last two days, starting with the three letters in 1992 which made it clear that all you were sharing, or considering sharing, or ever considered sharing, was royalties payable in respect of CAT's own patents, McCafferty and Winter II?

A.

I am sorry, this is your interpretation. I would not like to add anything more than I said. I cannot specifically recall all these details. What I can recall, I think, my Lord, I made, hopefully, clear to you and other people here.” (Day 8 p 81 et seq)

This suggestion that sharing downstream royalties was discussed with CAT and agreed was not made in Dr Müller-Neumann’s witness statement. There is no documentary evidence indicating that this ever occurred. It was neither made nor supported by any of Abbott’s other witnesses. It was not put to Dr Chiswell or Mr Nixon. If it had been true it would have been drawn explicitly to the attention of the Boards. It was not. I have come to the conclusion that it is one of the areas in which Dr Müller-Neumann’s evidence cannot be regarded as reliable.

In the result, there is no serious doubt as to the intention of the parties, their understanding of each other’s position and their mutual expression of their agreement on the topic of royalty sharing at the time the 1993 Agreement was entered into. It is as suggested by CAT. If, contrary to my findings above and on its proper construction, the Agreement does not accord with that common intention, it should be rectified so as to state expressly that the royalty sharing obligation applies only to third party rights which cover CAT’s library and phage display technology.

Since both parties agree that there was no change of intention between the 1993 Agreement and the 1995 one, the same conclusions apply to the latter also. Strictly it is not necessary to consider any more of the evidence. However I have already referred to Mr Shaw’s evidence relating to his putative rewording of Article 5.03. For reasons given already, this throws light both on his reliability as a witness and the intention and understanding of Knoll. There are one or two other documents which can be referred to from the post-1993 period which also throw light on the latter.

First is an internal memorandum dated 14 June 1995 sent by a Mr Anderson to Mr Shaw and copied to Dr Kamen and Dr Salfeld. Mr Anderson was an in-house attorney and later Director of Legal and Finance at BBC. He is still at Abbott but, unfortunately, was not produced to give evidence. In his document he was considering possible amendments to the 1993 Agreement which would be taken into the new 1995 Agreement. He said:

“A question has been raised pertaining to Section 5.03. The original agreement with CAT allows for a direct offset of the 2% royalty paid to the MRC for the Boss patent regarding antibody expression technology. The question is, if Boss patent is declared invalid and the Cabilly patent is in force in the United States, would it be possible to amend the wording of the CAT agreement to reflect this potentiality (I.e. offset for either Boss or Cabilly patents at the 2% level). Other royalties to third parties needed to practice the phage display technology are to be born equally between CAT and Knoll” (emphasis added)

The meaning of this is clear. Cabilly was not caught by the royalty sharing provisions, other patents, but only relating to the phage display technology, were. Mr Shaw did not respond to this. In the evidence before me he dismissed Mr Anderson as an accountant who was new to the law. I do not accept that evidence. It appears to me that Mr Anderson was simply recording what was the accepted understanding within Knoll of what was intended in the existing agreement between the parties.

On 30 June 1995 Mr Shaw wrote to Mr Nixon, copied to Dr Kamen and Dr Salfeld, and blind copied to, amongst others, Professor Schlick, Dr Dörper and Mr Anderson. With his letter he enclosed proposed amendments to the 1993 Agreement. As Mr Vos says, these included only immaterial amendments to Article 5.03 and none to 12.02. In his letter, Mr Shaw wrote:

“Although we have not proposed any modification to reflect this, we do wish to bring to your attention a concern we have with respect to Section 5.03. As you know, there is an interference involving MRC’s Boss patent and Genentech’s Cabilly patent. It is reasonable to assume that Genentech will prevail. If so, we are confronted with a 2% payment to MRC based on the Neuberger patent which is offset against the royalty due to CAT pursuant to Section 5.01, and a likely 3% royalty to Genentech which is not offset. We would appreciate CAT’s support in addressing this issue with MRC as well as an open discussion of how this might affect Section 5.03” (emphasis added)

Here, if ever, there was an opportunity to raise with CAT the question of sharing Cabilly royalties. Dr Chiswell said in his witness statement that he must have seen this letter at the time and that his understanding was that Cabilly did not fall within the royalty sharing provisions. He went on to say:

“141.

Knoll and CAT had been aware of the interference proceedings between the Boss patent and the Cabilly patent for some considerable time - they were referred to in Patent Confid and (as recorded in Knoll’s minutes of the meetings on 8 February 1993 (DJC1/101-104)) had been discussed in at least one of the meetings in 1993. At no time, so far as I was aware, over the course of the negotiations between CAT and Knoll did Knoll ever suggest that the Cabilly patent would or did fall within the scope of the royalty sharing provision in Article 5.03 or Article 12.02 of the 1993 Agreement ( or their corresponding provisions in the 1995 Agreement). Had Knoll done so or had I understood Bob Shaw's letter as contending that the Cabilly patent fell within the royalty sharing provisions, this would have raised a major issue. I would have had to report it to CAT's Board of Directors. Had Knoll persisted in an understanding of the Article which entailed the Cabilly patent falling within the royalty offset provisions, we would have had to open the discussion for the new agreement against the backdrop of Knoll contending Article 5.03 meant something different to the meaning which CAT understood Article 5.03 to have and which CAT understood Knoll to have agreed to in the negotiation of the 1993 Agreement.”

None of this was challenged in cross-examination. I have little doubt that, had Knoll been intending to assert that it was entitled under the 1993 Agreement to share Cabilly royalties this would have been stated clearly, would have been understood by Dr Chiswell and would have been soundly rejected. That, however, is not what the letter says. It contains no suggestion that there is an obligation to share Cabilly royalties. On the contrary, what it is saying is there is no such obligation. Mr Nixon’s evidence accords with Dr Chiswell’s. I accept it.

In relation to this, Mr Shaw gave evidence which I do not accept was truthful. He referred to the use of the word “offset” in the letter and suggested that all he was talking about was whether there was a 100% offset for Cabilly, as allowed for in respect of the MRC’s royalties in Article 5.03. He maintained that he still thought that Cabilly royalties would be shared 50/50. He wanted to move this up to 100%. He was using “offset” in this narrow way. I do not accept that he was using that word in such a limited way. If he were, it would be impossible to reconcile with Article 12.02 where the word “offset” is used in relation to exactly the same sharing provisions.

The letter speaks for itself. Contrary to the oral evidence he gave to me, Mr Shaw shared the same understanding as did Knoll itself. Cabilly was not covered by Article 5.03 at all. What was being proposed was a substantial change, namely extending the royalty burden on CAT to cover this patent as well.

The same point arises on a letter dated 14 July from Dr Salfeld to Professor Schlick. It concerns the possibility of Knoll taking licences under two patents; Enzon and Cabilly. It is not in dispute that Enzon was a patent which appeared to cover McCafferty’s phage display technology. As we know, Cabilly was much further down stream. If CAT is correct as to the understanding that not only it, but also Knoll, had as to the intended scope of Article 5.03, Enzon would fall within the royalty sharing obligation and Cabilly would not. That, in substance, is what this letter says. The relevant passages read as follows:

“We propose to take licenses under pending patents that we may or may not need for the antibody project in the future.

To Genentech: … In the US only an interference of [the Boss] patent with a Genentech patent (the ‘Cabilly patent’) has been pending for years. Because there has been no resolution yet – and we want to position ourself for any outcome prior to embarking on manufacturing – we propose to take a license under the ‘Cabilly patent’ for TNF and IL-12 this year, that would be effective once the interference has been resolved. The specifics are outlined in the attachment.

To Enzon: We have a complete set of licenses through CAT to all phage display patents owned by the MRC and CAT. The (sic) is one patent application in the area that predates the CAT applications. That patent is owned by Enzon and is called the ‘Ladner 1’ patent; … The BASF legal department also has good reason to believe that the Ladner 1 patent is unlikely to issue in the US in the current broad form.

We propose to position ourself again for any outcome and to take a license just in case. The royalties under such a patent are to be offset against royalties due CAT anyway.” (Emphasis added)

There can be no doubt, and Dr Salfeld did not dispute, that the word “offset” here refers to royalty sharing. Other than in this respect Dr Salfeld felt unable to answer any questions about this document. Once again, the meaning is clear. The Ladner patent was in the area of phage display. Royalties under “such a patent” were to be shared under the 1993 Agreement. This is to be contrasted with Cabilly. There is no suggestion that royalty sharing applied to it. This was Knoll’s understanding at that time and at all other relevant times.

There is one final matter which is worth referring to which supports CAT’s case. Axel was a patent which covers part of the post-phage process. It was referred to in a number of Knoll documents, only some of which are referred to above. It was never suggested to CAT by Knoll or Abbott, prior to these proceedings, that royalties under it fell to be shared. It is no different in that respect to any of the other patents Abbott now claims to be entitled to share royalties under, that is to say the patents listed in Schedules A and B to the Re-Amended Defence. I have commented already on the fact that it is most unlikely that Knoll would have failed to discuss this with CAT if, as it now says, it was always its understanding that CAT had agreed to such an obligation. I have also referred to the Cabilly patent which was discussed on a number of occasions with CAT. Not once during those conversations or correspondence was it suggested to CAT that it had an obligation to share the royalties due under that patent. On the contrary, when Knoll became agitated at the prospect of having to pay royalties to Genentech in respect of Cabilly it only asked CAT to intervene on its behalf with the MRC to try to persuade it to reduce the 2% royalty it was receiving for Boss and Neuberger. It is just not credible that it would have done that without, at the same time, mentioning that half the royalties were to be borne by CAT if that was its understand of what the parties had agreed. Axel is, in some respects, like Cabilly in the sense that the need to discuss royalty sharing with CAT – if there was any agreement to that effect - arose at an early stage. In fact Knoll entered into a licence under the Axel patent on 1 June 1995, that is to say before the 1995 Agreement was signed. Notwithstanding this, the claim to be entitled to share the royalty burden with CAT was not mentioned then or until this dispute arose in 2003, some eight years later.

I must emphasise that the above journey through the history of the negotiations between and actions of the parties has deliberately concentrated on a number of documents and events which appear to me to be particularly significant. Mr Vos relied on many more. He examined in detail the drafting of Articles 12.02 and the blow by blow negotiations designed to get the MRC to lower its royalty demands so as to reduce the burden on Knoll once it took a licence for Cabilly. He examined the attempts made by Knoll to pay the MRC directly, thereby taking its 2% out of the Agreements so that CAT would not obtain a “windfall”, as it put it, when the Boss and Neuberger patents expired or were not needed any more. I have not felt it necessary to go through all this and much other detail. It is sufficient to say that it all supports CAT’s case which is, in my view, overwhelming.

For these reasons CAT succeeds in these proceedings.