Hospira UK Ltd v Genentech, Inc. [2016] EWCA Civ 1185

Case No A3 2015 3238[2016] EWCA Civ 1185IN THE COURT OF APPEAL ( DIVISION)ON APPEAL FROM THE HIGH COURT OF JUSTICECHANCERY DIVISIONPATENTS COURTTHE HON MR JUSTICE ARNOLD[2015] EWHC 1796 (Pat)Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 30/11/2016
Before
LORD JUSTICE LONGMORELORD JUSTICE KITCHINLORD JUSTICE FLOYD
Between
HOSPIRA UK LIMITEDAppellantGENENTECH, INC.Respondent
Michael Tappin QC with Mark Chacksfield (instructed by Marks & Clerk Solicitors LLP) for AppellantRichard Meade QC, Thomas Mitcheson QC and Jeremy Heald (instructed by Taylor Wessing LLP) for RespondentHearing Hearing dates: 8 November 2016
Judgment
[1]The appellant, Genentech, Inc (“Genentech”), is the proprietor of European Patent (UK) No. 1037926 (“the patent”). The respondent, Hospira UK Ltd (“Hospira”) obtained an order for revocation of the patent before Arnold J, who, in his judgment dated 24 June 2015, found the claims of the patent to be novel but lacking in inventive step. Genentech appeals with permission granted by the judge himself.[2]The patent relates to the use of the antibody trastuzumab in combination with a chemotherapeutic agent, a taxane, for the treatment of HER2-positive breast cancer. At the priority date of the patent, 12 December 1997, a phase III clinical trial of trastuzumab in combination with a taxane (paclitaxel) was under way. The trial had already been described by Baselga et al in an article entitled “HER2 Overexpression and Paclitaxel Sensitivity in Breast Cancer: Therapeutic Implications”, Oncology, 11, Supp.2 43-48 (“Baselga 97”). Baselga 97 did not, however, disclose any results from the trial.[3]The patent summarises the results of the trial of trastuzumab in combination with paclitaxel described in Baselga 97. The results show that the combination treatment results in a substantial increase in time to disease progression (“TTP”) and tumour response rates in HER2-positive metastatic breast cancer patients compared to paclitaxel alone.[4]The results reported in the patent led to the approval of trastuzumab as a first line treatment for HER2 positive breast cancer, first in the US in 1998 and subsequently elsewhere. Trastuzumab, under its trade name Herceptin, has now become the standard of care worldwide for the majority of women with such cancers, usually given in combination with a taxane. It is a very successful drug.[5]Before coming to the issues on the appeal, it is necessary to say a little more about the patent itself and also about Baselga 97.

The patent

[6]At [0011], the patent explains that Herceptin “has been clinically active in patients with ErbB2-overexpressing metastatic breast cancers that had received extensive prior anti-cancer therapy.” ErbB2 and HER2 are alternative names for the same gene. At [0012] the specification notes that ErbB2 overexpression is commonly regarded as a predictor of a poor prognosis for the patient. It continues by explaining, perhaps somewhat paradoxically, that ErbB2 overexpression is associated with a greater clinical response to taxanes:
“However, despite the association of ErbB2 overexpression with poor prognosis, the odds of a HER2-positive patients responding clinically to treatment with taxanes were greater than three times those of HER2-negative patients… [Trastuzumab] was shown to enhance the activity of paclitaxel… against breast cancer xenografts in nude mice… which express high levels of HER-2.”
[7]Later, the specification sets out, by way of example, the clinical trial of the combined treatment. The patient population enrolled for the trial was split into two groups. One group was to receive trastuzumab and chemotherapy whilst the other was to receive only chemotherapeutic agents. For reasons explained, the chemotherapy administered was either paclitaxel or an anthracycline/cyclophosphamide treatment. The results of the clinical trial are then set out in the table which appears at [148]:[8]The specification says that the assessments of TTP and response rates showed a significant augmentation of the chemotherapeutic effect of paclitaxel, without an increase in overall severe adverse events.[9]The tabulated results show that, taking the two chemotherapy groups together, the addition of trastuzumab extended time to progression by 3.1 months and the response rate by 25.8%. Looking at each of the chemotherapy groups separately, there are increases in both time to progression and response rates as a result of the addition of trastuzumab. In the paclitaxel group, TTP was increased by 70% and response rate by 130%.[10]The specification concludes at [0150]:
“These data indicate that the combination of anti-ErbB2 antibody treatment with chemotherapy markedly increases the clinical benefit, as assessed by response rates and the evaluation of disease progression.”

The claims

[11]It is common ground that only claim one needs to be considered. This is in so-called Swiss form. The claim uses the term “taxoid”, but that is the same thing as a taxane. I adopt the judge’s breakdown of the claim into integers, which is as follows:
“[1] Use of an anti-ErbB2 antibody in the preparation of a medicament [2] for treatment to provide clinical benefit as measured by increased time to disease progression of malignant breast cancer characterised by overexpression of ErbB2 in a human patient, [3] wherein said antibody binds to epitope 4D5 within the ErbB2 extracellular domain sequence as determined by a cross-blocking assay using said antibody and antibody 4D5 obtainable from deposit ATCC CRL 10463, [4] and wherein the method comprises combined administration of the antibody with a chemotherapeutic agent which is a taxoid and not in combination with an anthracycline derivative, [5] wherein the combined administration has: [a] clinical efficacy as measured by determining time to disease progression and [b] reduced myocardial dysfunction compared with combined administration of the antibody and anthracycline derivatives.” [a] clinical efficacy as measured by determining time to disease progression and [b] reduced myocardial dysfunction compared with combined administration of the antibody and anthracycline derivatives.”

Baselga 97

[12]Baselga 97 is a review article. The authors were Jose Baselga, a member of the Oncology Service at the Hospital General Universitari, Val d’Hebron, Barcelona, Andrew Seidman and Larry Norton, of the Breast Medicine Service in the Department of Medicine at the Memorial Sloan-Kettering Cancer Center, New York, and Peter Rosen of the Department of Pathology at the same institution. The abstract reads as follows:
“Overexpression by the HER2 gene plays a significant role in breast cancer pathogenesis, and the phenomenon is commonly regarded as a predictor of a poor prognosis. HER2 overexpression has been linked to sensitivity and/or resistance to hormone therapy and chemotherapeutic regimens, including CMF (cyclophosphamide, methotrexate, and fluorouracil) and anthracyclines. Studies of patients with advanced disease demonstrate that, despite the association of HER2 overexpression with poor prognosis, the odds of HER2-positive patients responding clinically to taxanes were greater than three times those of HER2-negative patients. Further studies in preclinical models used combination therapy for breast cancer cells that overexpress HER2, and the use of agents that interfere with HER2 function plus paclitaxel (Taxol) resulted in significant antitumor effects.”
[13]The introductory section of Baselga 97 explains that the taxanes are an important new class of anti-cancer agents with a unique mechanism of action. Paclitaxel was selected for clinical development based on impressive anti-tumour activity in xenograft tests. It goes on to explain that paclitaxel has been shown to have a high degree of anti-tumour activity in women with metastatic breast cancer. The introductory section concludes:
“An area of increasing interest in clinical research on taxanes is the possible role of oncogenes, such as HER2, in determining clinical response to paclitaxel. Studies have examined whether strategies can be designed to increase the agent's efficacy (or curb resistance to it) in breast cancers that overexpress HER2. Available data that will be presented in this review suggest that HER2 overexpression may influence the response to paclitaxel in patients with metastatic breast cancer and that anti-HER2 monoclonal antibodies significantly increase the antitumor activity of paclitaxel in vitro and in vivo.”
[14]The suggestion that an anti-HER2 antibody (such as trastuzumab) may enhance the anti-tumour activity of the taxane paclitaxel in vivo is close to a teaching that the two should be administered in combination.[15]Under the heading “Role of Monoclonal Antibodies”, Baselga 97 explains that the murine monoclonal antibody 4D5 directed against the extracellular domain of p185HER2 had been found to be a potent inhibitor of human breast cancer cells overexpressing HER2 in mouse xenograft studies. These studies, it is reported, led to the development of a humanised anti-p185HER2 antibody, namely trastuzumab. Trastuzumab was:
“found to be safe and to have dose-dependent pharmacokinetics in two prior phase I clinical trials.”
[16]Another section of Baselga 97 is headed “HER2 overexpression and taxane sensitivity”. This section reports work at the Memorial Sloan-Kettering Cancer Center investigating the possible relationship between HER2 overexpression and response to taxanes in patients with metastatic breast cancer. The authors report that:
“In 37.7% of patients, tumors were positive by immunohistochemistry with the 4D5 antibody. The overall response to taxanes for all patients in this analysis was 46.7%. Remarkably, 65.2% of patients with HER2-positive tumors responded vs 35.5% of patients with HER2-negative tumors ... Thus, despite a positive correlation of HER2 expression and poor prognostic features, the odds of HER2-positive patients responding clinically to taxanes were greater than 3 times those of HER2-negative patients.”
[17]A further section of particular relevance is headed “Combined Therapy with Anti-HER2 Agents and Chemotherapy”. The authors say:
“A novel approach that is currently being explored is the combined use of therapies directed at p185HER2, such as MoAbs given alone or in combination with conventional chemotherapeutic agents, including paclitaxel.”
[18]The authors then go on to explain research conducted on antibodies against p185HER2, in particular studies at Genentech, Inc. They explain that in a phase II clinical study with trastuzumab in patients who had prior extensive metastatic disease and treatment, toxicity was minimal, and there was an overall response rate of 11.6%. “Thus, [trastuzumab] is clinically active in patients who have metastatic breast cancers that overexpress HER2 and have received extensive prior therapy. A confirmatory study that will include 200 patients with less heavily pretreated metastatic disease is currently underway.”[19]The authors then go on to explain that in order to enhance the anti-tumour activity of these monoclonal antibodies, several investigators have used them in combination with chemotherapeutic agents. Amongst these, they refer to work by Slamon who had demonstrated that the murine antibody 4D5 enhanced sensitivity to cisplatin in cisplatin-resistant ovarian carcinoma cell lines. Cisplatin is a different chemotherapeutic agent, and not a taxane.[20]A further section then discusses the activity of the chemotherapeutic agents paclitaxel and doxorubicin in combination with trastuzumab. These studies were conducted in monolayer cell culture in soft agar as well as with breast cancer human tumour xenografts in mice. The authors say:
“Therapy with MoAb 4D5 alone produced a 35% growth inhibition, and paclitaxel alone resulted in a 35% growth inhibition when compared with animals treated with a control MoAb. The treatment with paclitaxel plus 4D5 resulted in major antitumor activity, with 93% inhibition of growth. This result was markedly better than an equipotent dose of doxorubicin … and 4D5… In addition, paclitaxel combined with 4D5 resulted in the disappearance of well-established xenografts.”
[21]The authors also report on a phase II study of trastuzumab in combination with cisplatin in patients with breast carcinomas that overexpress p185HER2 with a history of proven resistance to chemotherapy. The authors say:
“In this group of patients with expected cisplatin resistance, the observed response rate to the combined therapy was 25%, suggesting that the synergy observed in the laboratory was reproducible in the clinic. In addition the combined therapy was no more toxic than cisplatin alone.”
[22]The authors then go on to describe the ongoing phase III study of trastuzumab in combination with chemotherapeutic agents. They begin by saying that the results from the phase II studies and the activity of trastuzumab against xenografts when given in combination with doxorubicin and paclitaxel were “encouraging”. Those positive results had led to the design of a phase III multinational study of chemotherapy in combination with trastuzumab in patients with HER2-overexpressing breast tumours who have not received prior chemotherapy to metastatic disease. Figure 2 sets out the plan of the phase III clinical trial showing two arms, one with only chemotherapy and the other with chemotherapy plus trastuzumab. The endpoint is said to be time to disease progression. The main goal of the study was to determine whether the addition of the anti-HER2 antibody increases the time to disease progression compared with the group of patients treated with chemotherapy alone.[23]The overall conclusion of the article is this:
“The predictive value of HER2 overexpression for paclitaxel response requires independent confirmation in advanced disease and early breast cancer. This question could be addressed in ongoing studies of paclitaxel-based adjuvant therapy in breast cancer. In preclinical models, the combined therapy of breast cancer cells that overexpress HER2 with agents that interfere with HER2 function and paclitaxel results in a marked antitumor effect. One such strategy, the combination of anti-HER2 MoAb with paclitaxel, is currently being evaluated. If the results of these studies are positive, we might be faced with a novel paradox, in which expression of a receptor that confers a worse prognosis provides us with an opportunity for increased response to taxanes.”

Obviousness – law

[24]The law of obviousness, and the relevance of “obvious to try” has been addressed by a number of recent judgments in this court. I attempted to summarise the effect of these decisions, with the agreement of Kitchin and David Richards LJJ, in an earlier case between these parties, Hospira UK Limited v Genentech, Inc [2016] EWCA Civ 780 at [9] to [14]:
“9. I start with the well-known passage from the judgment of Kitchin J (as he then was) in Generics (UK) Ltd v H. Lundbeck A/S [2007] EW HC 1040 (Pat), at [72], which has been approved at the highest level: “The question of obviousness must be considered on the facts of each case. The court must consider the weight to be attached to any particular factor in the light of all the relevant circumstances. These may include such matters as the motive to find a solution to the problem the patent addresses, the number and extent of the possible avenues of research, the effort involved in pursuing them and the expectation of success.” 10. It follows that the court is required to embark on a multi-factorial assessment. The approach of the appellate court to such questions is well-known: the decision is not open to independent evaluation in this court unless the judge has made an error of principle: see Biogen Inc v Medeva Plc [1997] RPC 1 at [45] per Lord Hoffmann. 11. It is also well-settled that "obvious to try" is not a substitute test for obviousness: see e.g. per Lewison LJ in Medimmune Ltd v Novartis Pharmaceuticals UK Ltd [2012] EWCA Civ 1234, [2013] RPC 27 at [181]. There is only one statutory question, namely whether the invention was obvious at the priority date. In the same case at [90] to [95] Kitchin LJ explained that whether the invention was obvious to try was merely one of many considerations which it may be appropriate for the court to take into account in addressing the statutory question. It must in any case be coupled with a reasonable or fair prospect of success. At [91] Kitchin LJ emphasised: “Whether a route has a reasonable or fair prospect of success will depend upon all the circumstances including an ability rationally to predict a successful outcome, how long the project may take, the extent to which the field is unexplored, the complexity or otherwise of any necessary experiments, whether such experiments can be performed by routine means and whether the skilled person will have to make a series of correct decisions along the way.” 12. Thus a judge's assessment whether an approach has a reasonable or fair prospect of success is itself another multi-factorial assessment, where this court should again pay proper respect to the evaluation of a trial judge. 13. There is also no single standard of what amounts to a fair expectation of success. Thus Jacob LJ's statement in Saint Gobain PAM SA v Fusion-Provida Ltd [2005] EWCA Civ 177 at [35] that it must be "more-or-less self-evident that what is being tested ought to work" is far from being a test of universal application. In Novartis AG v Generics (UK) Ltd (trading as Mylan) [2012] EWCA Civ 1623 Kitchin LJ (with whom Lewison and Munby LJJ agreed) warned against imposing a straitjacket on the law by adopting any form of words as a standard. He said at 55: “What is a reasonable or fair expectation of success will again depend upon all the circumstances and will vary from case to case. Sometimes, as in Saint Gobain, it may be appropriate to consider whether it is more or less self-evident that what is being tested ought to work. So, as this court explained in that case, simply including something in a research project in the hope that something might turn up is unlikely to be enough. But I reject the submission that the court can only make a finding of obviousness where it is manifest that the test ought to work. That would be to impose a straitjacket on the assessment of obviousness which is not warranted by the statutory test and would, for example, preclude a finding of obviousness in the case where the results of an entirely routine test are unpredictable.” 14. Indeed, so much was confirmed by Lord Hoffmann in Conor Medsystems v Angiotech Pharmaceuticals [2008] UKHL 49, [2008] RPC 28 at [42], approving Jacob LJ's comprehensive review of the authorities in this area in the Court of Appeal in that case, including his view that “how much of an expectation depended upon the particular facts of the case.”” “The question of obviousness must be considered on the facts of each case. The court must consider the weight to be attached to any particular factor in the light of all the relevant circumstances. These may include such matters as the motive to find a solution to the problem the patent addresses, the number and extent of the possible avenues of research, the effort involved in pursuing them and the expectation of success.” “Whether a route has a reasonable or fair prospect of success will depend upon all the circumstances including an ability rationally to predict a successful outcome, how long the project may take, the extent to which the field is unexplored, the complexity or otherwise of any necessary experiments, whether such experiments can be performed by routine means and whether the skilled person will have to make a series of correct decisions along the way.” “What is a reasonable or fair expectation of success will again depend upon all the circumstances and will vary from case to case. Sometimes, as in Saint Gobain, it may be appropriate to consider whether it is more or less self-evident that what is being tested ought to work. So, as this court explained in that case, simply including something in a research project in the hope that something might turn up is unlikely to be enough. But I reject the submission that the court can only make a finding of obviousness where it is manifest that the test ought to work. That would be to impose a straitjacket on the assessment of obviousness which is not warranted by the statutory test and would, for example, preclude a finding of obviousness in the case where the results of an entirely routine test are unpredictable.”

The judgment of Arnold J

[25]The judge commenced his assessment of the issue of inventive step at paragraph 116 of his judgment. However it is worth noting that, in construing the claim, he had recorded some important points which reflect the correct target for the obviousness attack. Two are important here.[26]Firstly, it was common ground that the requirements for “clinical benefit” in integer [2] and “clinical efficacy” in integer [5][a] of claim 1 were functional technical features of the claim in the sense that the actual attaining of the clinical benefit was a requirement of the claimed invention. That conclusion reflects the decision of this court in Regeneron Pharmaceuticals Inc v Genentech Inc [2013] EWCA Civ 93 at [56].[27]Secondly it was common ground that the word “for” in integer [2] imports a mental element, namely that it is known to, or reasonably foreseeable by, the manufacturer of the medicament that the antibody will be intentionally administered in combination with a taxane for the relevant therapeutic purpose. The judge held that this meant that the person administering the combination must intend to achieve the clinical benefit required by the claim, that is increased efficacy as compared to the taxane alone. The common ground reflected this court’s decision in Warner-Lambert Company LLC v Actavis Group PTC EHF [2015] EWCA Civ 556.[28]It was on the basis of this construction of the claim that the judge rejected Hospira’s attack of lack of novelty on claim 1. Applying Lord Hoffmann’s well known analysis of the law in Synthon BV v SmithKline Beecham plc [2005] UKHL 59; [2006] RPC 10, Hospira had argued that the inevitable result of carrying out the teaching of Baselga 97, notably the described and ongoing clinical trial, would result in infringement of the claim. However the judge rejected the novelty attack on the basis the person administering the combination would not have the mental element required by the claim, because he or she could not intend to administer the combination to achieve the clinical benefit unless he or she knew that that clinical benefit would be achieved. The judge’s finding to that effect is not in issue on this appeal.[29]Turning to the assessment of obviousness the judge said at [116] that he did not understand it to be in dispute that it would not be obvious to the skilled person from reading Baselga 97 that the combination would result in increased efficacy as measured by TTP compared to taxane alone. He continued that it was also not in dispute that the skilled person would concur with the organisers of the phase III trial that such a trial was worth undertaking at least because of the hope that it would succeed and the benefit that would be obtained if it did. Thus, for the judge: “The issue is whether the skilled person would have a fair expectation of success, meaning a fair expectation of increased efficacy as measured by TTP of the combination compared to taxanes alone if he were to undertake such a trial”.[30]The judge then said that in resolving this issue the key question was how the skilled person would interpret Baselga 97, and in particular as to how encouraging the skilled reader would consider the results from previous studies reported in Baselga 97 to be:
“Putting it simplistically, would the skilled person think that the glass was half full (or more than half full, as Hospira contend) or half empty (or more than half empty as Genentech contend)?”
[31]Before addressing that question directly, however, the judge conducted a systematic analysis of the relevant circumstances: i) He first considered motivation, concluding that the skilled person would be strongly motivated to find an effective treatment for HER2-positive breast cancer, which was badly needed. ii) Next the judge tackled whether the trial would be a routine one. He said that it was common ground that phase III trials were regarded as routine in the sense that they were well-known and frequently undertaken. Moreover in the present case that was particularly so, given that the design of the trial had already been established and described in Baselga 97. iii) Nevertheless, it was common ground that phase III trials were regarded as logistically complicated, time-consuming and expensive. That, however, did not mean that the trial described in Baselga 97 would be regarded as technically burdensome. The trials would be no more technically burdensome than carrying out the teaching of the patent itself. They would thus take time and money, but would not be technically burdensome. iv) Next, the judge considered whether the trial would present a risk to patients. He pointed out that the skilled person would appreciate that trastuzumab, as a humanised antibody, was designed to present a low risk of side effects, and that Baselga 97 presented evidence that this was indeed the case. Moreover, as all the patients in the trial would receive standard chemotherapy, they would not miss out on a treatment they would otherwise receive. v) The judge also summarised what he considered to be the alternative options for the skilled person to follow. These were(a) taking monotherapy with trastuzumab further prior to combining the therapies,(b) carrying out a phase II trial of the combination before carrying out the phase III trial and(c) following up the anthracycline limb of the phase III trial. He considered these options to be very limited and none of them to detract from the obviousness of carrying out the phase III trial of trastuzumab and taxane. vi) Next the judge considered what he called the scientific context. This included (a) the general rationale for combined therapies and (b) the prior work on trastuzumab and combinations thereof. As to (a), he held that the skilled person would know that combined therapy was a well-established approach to the treatment of cancer and that the combination of trastuzumab and a taxane fitted the rationale for combining agents. This conclusion refers back to an earlier section in the judgment at paragraph 37 where the judge explained the rationale in the following terms: “The evaluation and use of combination therapy, i.e. administering combinations of different drugs, was common. The idea behind combination therapy is to kill the cancer cells to the greatest extent possible in a single course of treatment by using different agents with differing methods of action and minimal overlap in toxicity profiles. In some instances combination therapy turned out not to provide any additional benefit, but these examples did not detract from the general understanding that a logical combination of two drugs which were individually effective, which were directed at different targets and which did not have significantly overlapping toxicity would be expected to have greater efficacy than either drug individually. It was standard practice to start to experiment with combinations even before a licence was obtained for monotherapy with a new agent.” vii) Under the heading “No lions in the path”, the judge dealt with an argument by Genentech that, as taxanes were among the leading chemotherapeutic agents available, it was a “big ask” to try and obtain any increased efficacy as compared with the taxane alone. The judge accepted this, but did not think it detracted from the message conveyed by Baselga 97. viii) Finally in this section the judge dealt with failure rates in phase III trials. The judge accepted that the skilled person would know that it was not at all uncommon for phase III trials to fail and would appreciate that the fact that a phase III was being undertaken would not necessarily mean that it would be successful.[32]The judge then turned to how the skilled person would view the results reported in Baselga 97. So far as the results of the Phase II trial of trastuzumab monotherapy were concerned, the parties disagreed as to how these would be viewed. Prof Calvert thought that the response rate of 11.6% would be seen as promising because the patients had been more heavily pre-treated than those who participated in early trials of doxorubicin and paclitaxel and because trastuzumab was a novel agent with a novel mechanism, yet it was producing results in a patient group which it was difficult to get responses from. Notwithstanding this, the judge concluded that the results would be seen as disappointing, but enough to take the drug (i.e. the monotherapy) further into clinical trials, accepting the evidence of Prof Barrett-Lee.[33]As for the results on trastuzumab with cisplatin, the dispute again centred on the response rate, in this case 25%. On these results Prof Calvert’s view was that this was promising, since the expected response rate from pre-treated cisplatin patients would have been 10%. Prof Barrett-Lee did not agree that it was promising, only that it was worth pursuing further. The judge held that the skilled person would regard this result as “a small additional cause for optimism.”[34]So far as the mouse xenograft studies are concerned, the dispute centred on how predictive the skilled reader would consider these to be of efficacy in humans. The judge had already reached a conclusion on this issue, at paragraphs [39] to [40]. He said:
“39. There is disagreement between the parties as to how reliable positive results in mouse xenograft studies were perceived to be as predictors of efficacy in humans. Genentech contend that mouse xenograft data were recognised to be unreliable, whereas Hospira contend that they were universally used to judge whether to progress a drug to clinical trials and were a good model for testing drug interactions. In my judgment the evidence demonstrates that both parties are correct. 40. As Prof Barrett-Lee explained, mouse xenograft models had a number of well-recognised shortcomings, in particular the fact that the rest of the biological system was murine and that the mice did not have functioning immune systems. As a result, as he put it, "it was well-known that mouse xenografts were not always reliable predictors of clinical efficacy in humans [my emphasis]". Accordingly, he said, "evidence of activity [in mouse xenografts] provided little assurance that the agent would ultimately demonstrate clinical efficacy in humans [my emphasis]". Prof Calvert agreed that mouse xenograft models did not always predict efficacy accurately. He insisted, however, that they were universally used to decide whether to take drugs forward. Prof Calvert also said that they were a good model for testing drug interactions. I did not understand Prof Barrett-Lee to dispute these points, subject to his overriding point that mouse xenograft models were not considered to be reliable predictors of clinical efficacy in humans.”
[35]So far as the mouse xenograft results reported in Baselga 97 were concerned, Prof Calvert's view was that the results for the combination of trastuzumab and paclitaxel were striking, because they showed 93% inhibition of growth compared to 35% for either agent alone, and the disappearance of well-established xenografts. Prof Barrett-Lee also agreed that the xenograft data was "very striking", but maintained that the skilled person would not draw any conclusions from it regarding humans. The judge’s view was that the skilled person would be aware that the xenograft results were not reliably predictive of what would happen in humans, but would nevertheless regard them as promising.[36]Overall, the judge considered that the skilled person would agree with the assessment of the authors of Baselga 97, namely that results of the previous studies were "positive" and "encouraging" and justified the decision to move forward to a Phase III trial even though there had been no Phase II trial of the combination of trastuzumab and a taxane.

The judge expressed his conclusion at [134] as follows:

“Prof Calvert’s overall view was that the skilled person would have a reasonable expectation of success in the Phase III trial. Prof Barrett-Lee was not prepared to go that far, but he accepted that it was justified to proceed with the Phase III trial. In my judgment the conclusion to be drawn from the matters considered above is that the skilled person would have had a fair expectation of success in the sense defined above. Accordingly the claimed inventions were obvious.”

The appeal

[38]Mr Tappin QC, who presented the appeal on behalf of Genentech with Mr Mark Chacksfield, contended that the judge had fallen into two major errors in his analysis. The first was that he had taken a wrong approach to what was a fair expectation of success. The second was that the judge had failed to take proper account of the position of the skilled person.[39]On the first of these alleged errors, Mr Tappin submitted that the judge had assessed whether there was a fair expectation of success by considering only whether the existing results ‘justified’ Genentech’s decision to proceed with the phase III trial. That was a flawed test, he submitted. The judge should have asked himself, following Saint Gobain (supra) whether it was “more or less self-evident that what is being tested ought to work”.[40]There are, as it seems to me, two parts to Mr Tappin’s submission. The first is what is meant by the test of “a fair expectation of success” in a case such as this. The second is whether the judge correctly gave effect to that test.[41]As to the first of these points, Mr Tappin submitted that where the claim was, as here, to the use of a drug or drug combination to achieve a claimed clinical effect, it was necessary for the court to focus on the technical subject matter of the claim. In the present case the technical subject matter of the claim included the actual attaining of the clinical benefit, as the judge had held. Mr Tappin developed this point by reference to some passages in Conor v Angiotech (supra). In that case the claim was to a coronary stent coated with taxol for the prevention of restenosis. Because the specification had, in the trial judge’s view, not established that the claimed product would in fact prevent restenosis, he had approached obviousness of the basis of whether it was obvious to include a taxol-coated stent in a trial to see whether it would work. At [17] of his speech in the House of Lords, Lord Hoffmann rejected that approach as:
“..an illegitimate amalgam of the requirements of inventiveness (Art. 56 of the EPC) and either sufficiency (Art. 83) or support (Art. 84) or both. It is the claimed invention which has to involve an inventive step. … The alleged inventiveness lay in the claim that the product would have a particular property, namely to prevent or treat restenosis. … So the question of obviousness was whether it was obvious to use a taxol-coated stent for this purpose.”
[42]At paragraphs [27] and[28] Lord Hoffmann said:
“In my opinion it is absolutely clear that the teaching of the specification, so far as it supported claim 12, was that a taxol-coated stent would prevent or treat restenosis. … The question was whether that was obvious and not whether it was obvious that taxol (among many other products) might have this effect.”
[43]Putting these two sentences together, Mr Tappin submits that the sole question was whether it was obvious that a taxol-coated stent would prevent or treat restenosis. However, as Mr Richard Meade QC, who appeared for Hospira with Mr Thomas Mitcheson QC and Mr Jeremy Heald, pointed out, a rigid application of such a principle would leave no room for treating an invention as obvious when it was obvious to try. That cannot have been Lord Hoffmann’s intention, because in paragraph [42] he expressly approved Jacob LJ’s summary (in the Court of Appeal in that case) of the authorities on obvious to try, expressly recognising that the concept was useful where there was a fair expectation of success.[44]Recognising this problem, Mr Tappin stopped short of submitting that it must be shown that it was obvious that the claimed combination would (i.e. necessarily would) have the claimed clinical effect. He submitted, instead, that in such cases it was necessary for there to be a very high expectation of success, which he equated with the Saint Gobain standard – more or less self- evident that it ought to work.[45]I cannot accept even this more limited submission. It amounts to the creation of a lex specialis for claims which include as part of their technical subject matter a therapeutic effect or benefit. I can see no basis for imposing such a rigid rule if the decision in Conor does not mandate it (which it does not). The test is the flexible one described in the passages which I have cited above when dealing with the law, based on asking whether there is in all the relevant circumstances a fair expectation of success.[46]It is true that the claim in the present case did include, as part of its technical subject matter, that the claimed therapeutic benefit is actually achieved. It is very common for claims, whether expressly or by necessary implication, to include a feature which secures a defined technical effect. Such claims may nevertheless be obvious if it is obvious to try to achieve that technical effect by making something within the claims, provided that there is the necessary fair prospect of success.[47]The second part of Mr Tappin’s submission on this part of the case focuses on an important distinction. In pharmaceutical research a company may often have a strong motivation to explore avenues with little or perhaps even no expectation of useful results. Such a company may regard the exploration of such avenues as “justified” because the rewards are great if the research alights, against all the odds, on a successful treatment. The inference cannot be drawn, however, from the mere fact that such a company has embarked on such work, or regarded it as justified, that it did so with the required reasonable or fair expectation of success. That is, at least in part, because the evaluation of expectation of success includes the strength of the scientific reasoning which would predict a successful outcome. It would be a serious error to treat an invention as obvious simply on the basis of evidence that a clinical scientist would regard a trial as justified.[48]Based on this distinction, Mr Tappin submits that the judge fell into precisely this error. He draws attention to the judge’s intermediate conclusions in relation to the various components of the research reported in Baselga 97. Thus the trastuzumab monotherapy results were “disappointing” but “sufficiently encouraging to justify progression to further clinical trials”. The phase II trastuzumab/cisplatin trials were only “a small additional cause for optimism”. The xenograft studies were “promising”, albeit with a recognition that such studies were not reliably predictive. The judge’s overall conclusion in paragraph 131 went no further than a finding that the evidence justified a decision to move forward to the Phase III trial.[49]I do not think the judge fell into the error for which Mr Tappin contends. In expressing his final conclusion in paragraph 134 the judge expressly distinguished between Prof Calvert’s view that there was “a reasonable prospect of success in the Phase III trial” and Prof Barrett-Lee’s view which he said “did not go that far”, merely accepting that it was justified to proceed with the Phase III trial. In concluding that there was a fair prospect of success the judge was plainly preferring Prof Calvert’s evidence, which was based on the correct test, to that of Prof Barrett-Lee, which was not.[50]Prof Calvert’s view was plainly one which the judge was entitled to accept. As Mr Meade pointed out, the judge had found that combination therapy was common, and that these two drugs fitted the rationale for combination therapy. Trastuzumab and paclitaxel were drugs which were directed at different targets, having different modes of action. Trastuzumab had been shown to be free of significant side effects. Baselga 97’s proposal for a combination therapy would therefore have struck the skilled person as “a logical combination of two drugs directed at different targets and which did not have significantly overlapping toxicity”. The judge found as a fact that such combinations would be expected to have greater efficacy than either drug individually. It was also common practice to start to experiment with combinations even before a licence was obtained for monotherapy with a new agent. Against that, Genentech had put forward no convincing reason why the skilled person would think that the combination would not yield the relevant clinical benefit – there were no lions in the path.[51]In connection with this last point, it is true that on a number of occasions Prof Barrett-Lee had said that it was a “big ask” to improve on the admittedly excellent therapeutic properties of paclitaxel. But it is clear that the judge weighed that evidence against the unusually strong scientific rationale that the combination would produce a clinical benefit. Given the different targets and different modes of actions, there was no reason why a significant improvement could not be achieved.[52]It follows that, in reaching his findings, the judge was relying on much more than a mere inference from the fact that a pharmaceutical company considered the phase III trial to be justified. Baselga 97, and the expert evidence given in relation to it, plainly formed a sound basis on which the judge could find a fair expectation of success.[53]On the second alleged error, that the judge had failed to place himself in the position of the skilled person, Mr Tappin pointed out that the skilled person who is given Baselga 97 does not have trastuzumab or another anti-HER2 antibody in hand. In order to carry out a phase III trial of such an antibody in combination with paclitaxel the skilled person would have to make and purify an antibody, to formulate it for administration to humans, and then would have to carry out preclinical and early clinical trials to check its activity and safety. Mr Tappin drew our attention to some evidence given by Prof Barrett-Lee to the effect that the burden of that work would be something that the skilled person would have to balance against the uncertainties surrounding a successful outcome. Prof Barrett-Lee said that a skilled person, even one with a large organisation behind them, would have considered “the stakes too high and the odds too long”. Prof Calvert had agreed that one would have to be sufficiently convinced of the outcome to embark on such a course of work.[54]Mr Tappin submitted that the judge had wrongly dismissed the need for anyone other than Genentech to obtain and purify the antibody and perform Phase I and Phase II clinical trials. It was no answer to say, as the judge had, that the patent itself assumed that these were steps which the skilled person was capable of undertaking, and to suggest that it would be invalid for insufficiency if these steps imposed an undue burden on the skilled person. The relevance of the burdensome nature of these steps was that they fell to be weighed against the expectation of success, in order to reach an overall conclusion on whether it was obvious to try. The reader of the patent was in a different position, knowing that the clinical benefit would be achieved. The point which the judge never grappled with was whether the skilled person would embark on this burdensome work when he only had the level of expectation which Baselga 97 provides.[55]I am not at all persuaded by this submission. It is important to have in mind what

the judge actually said about the burden in paragraph 120:

“Would the trial be burdensome? It is common ground that Phase III trials were regarded as logistically complicated, time-consuming and expensive. But that does not mean that the Phase III trial described in Baselga 97 would be regarded as technically burdensome. Counsel for Genentech submitted that the trial would be burdensome because of the need for anyone other than Genentech to make, purify and formulate an anti-HER2 antibody and conduct pre-clinical and clinical trials on it before starting the Phase III trial. (Rightly, he did not rely upon the fact that no one else could use trastuzumab itself because of Genentech's basic patent.) As counsel for Hospira pointed out, however, this is no more than the skilled person would have to do in order to implement the Patent. If the skilled person could not do these things without undue burden, the Patent would be insufficient. Although the Patent contains instructions on how to do these things, it is not suggested that those assisting the skilled person would not know how to do them without such instructions, nor is it suggested that the instructions would reduce the amount of work required. Thus the work would take time and money, but it would not be technically burdensome.”
[56]Thus the judge is not saying that the trials would not involve any burden. He is accepting that there would be significant time and money tied up in getting to the point where one could carry out the trial, and in carrying it out. He points out, however, that the work is, in technical terms, within the capacity of the skilled person, because it is no more than the patentee expected of the skilled person when faced with carrying out the teaching of the patent. He was entitled to use the patent specification as a guide to the nature and burden of the work involved.[57]The judge plainly had the nature and burden of the work in mind in making his assessment of whether there was a fair expectation of success. It was not necessary for him, in expressing his conclusions in paragraph 134, to rehearse again all the factors which he had considered at earlier parts of his judgment. The judge sufficiently set out the factors which he had taken into account in his evaluation, and there is no basis for thinking that he had not factored in the time and cost involved in the trials.[58]Mr Tappin had a final point based on the judge’s treatment of the secondary evidence. Put in simple terms it was based on a contrast between the interest expressed in the publication of the individual trial reports (eventually drawn together in Baselga 97), and the publication of the results of the work reported in the patent. At the ASCO meeting in 1995 the results of the phase II trastuzumab monotherapy were presented, as well as those for the phase II trastuzumab/cisplatin trial. ASCO is the major meeting for clinical oncology in the world, attended that year by some 10,000 oncologists. The judge said that the publication of these results did not generate much attention. Subsequently the phase II monotherapy results and the mouse xenograft results were also published in the Journal of Clinical Oncology in 1996. The Journal of Clinical Oncology is the major journal in the field, read according to the evidence by some tens of thousands of oncologists. The judge found, again, that this did not generate attention.[59]By contrast, when the results reported in the patent were made known at ASCO 1998, some 5000 people attended to hear the presentation, having been alerted in advance by the abstract. The judge found the results generated “considerable excitement”. However, the judge was unimpressed:
“… I do not consider that the excitement generated by the presentation of the Phase III results at ASCO 1998 assists Genentech. The excitement was generated by the confirmation of a new and more efficacious therapy for HER2-positive breast cancer. That represented an important advance, but it does not show that the skilled reader of Baselga 97 would not have had a fair expectation of success. That is particularly true if, as is implicit in Genentech's case, those who were excited at ASCO 1998 were not familiar with Baselga 96 (in which case they would unlikely to have been familiar with Baselga 94 either).”
[60]I agree with the judge that the contrast in the reception given to the two sets of results does not throw much light on the issue he had to decide. What matters to clinical oncologists is the successful treatment of patients. The earlier results, whilst promising, had not got that far. Moreover the results reported at ASCO 98 were, quantitatively, very striking. It is necessary to bear in mind that the claims of the patent are more extensive, extending to any increase in time to progression, however small. I am less convinced that the judge could realistically assume that the audience was not sufficiently aware of the earlier results, but, in the circumstances, that does not matter.

Conclusion

[61]Overall, I do not think Arnold J’s analysis displays any of the errors of principle for which Mr Tappin contends. For the reasons I have given, I would dismiss the appeal.[62]I agree.[63]I also agree.