Last week saw the revocation of Bristol-Myers Squibb’s (BMS) European patent no. 1169038, which claimed the $1.6 billion a year anti-cancer drug dasatinib, aka Sprycel®. GJE partner, Arnie Clarke, who was representing an opponent in these proceedings, explains the decision and its potentially far-reaching consequences.
Background and Issues
The patent had been amended during the proceedings such that a single claim directed to a single compound was presented. Received wisdom was that such a narrow patent, directed to a successfully commercialised new chemical entity, must be patentable.
This case has created waves in the pharmaceutical industry, with representatives of innovator and generics companies alike showing a keen interest in the outcome. The hearing was attended by as many bloggers and journalists as professional representatives, and even the European Federation of Pharmaceutical Industries and Associations (EFPIA) got in on the act, filing observations at the EPO during the written proceedings.
The patent application as filed is a classic new chemical entity (NCE) case, with hundreds of exemplified and characterised compounds. One of these turned out to be dasatinib, although it was in no way singled out in the myriad of compounds disclosed.
The specification also claimed that the compounds of the invention exhibited protein kinase inhibition, and set out a list of assays said to be suitable for the assessment of this activity. It was also stated in the specification that:
“Compounds described in the following Examples have been tested in one or more of these assays, and have shown activity.”
However, no activity data was provided in respect of any of the compounds, and the patent was revoked on the basis of a lack of inventive step, essentially due to a lack of a verifiable technical effect.
On the face of it, this seems rather strange as the patent clearly disclosed dasatinib along with its intended use and associated pharmacological activity. While no activity data was presented in the application as filed, supporting evidence of protein kinase inhibition was filed during prosecution of the application in the EPO, but after the filing date of the patent. Such post-filed data is commonly accepted by the EPO in support of a technical effect disclosed in the patent application. However, the application as filed must have been deemed to make the technical effect credible before such post-filed data can be used to confirm or extrapolate the effect.
In the present case, as the application was deemed not to make the technical effect plausible, the post-filed data could not be used to supplement the original teaching. Consequently, the problem to be solved by the patent was found not to be the provision of protein kinase inhibitors, but merely the provision of further organic compounds. As such, a finding of lack of inventive step was inevitable.
This decision effectively raises the threshold level of the information content of a patent specification, and has significant implications on the patentability of pharmaceutical and life-sciences inventions.
The Opposition and Appeal Proceedings
During the granting and opposition proceedings at the EPO, the claims had been narrowed down to a single claim to a single compound – dasatinib.
Novelty was not in issue, and although the sufficiency of disclosure has been an issue in the first instance proceedings, the patentee had resolved this through amendment. In terms of inventive step, the prior art was quite structurally remote from dasatinib. However, the main issue was whether the application as filed rendered the claimed technical effect plausible, per se, such that supporting data, filed by BMS during examination of the application, could be used to ‘confirm’ the claimed effect.
In addition to the assertions made in the patent application, the patentee filed various expert witness testimonies arguing that the effect was plausible based on a structure activity relationship analysis. BMS also filed a declaration by an inventor who stated that BMS had made and tested the activity of the compounds, all of which showed activity. However, none of this was independently verifiable, and effectively amounted to a plea that the public should accept that what was stated in the application was the truth.
Although the written decision has not been rendered yet, the Board of Appeal appears to have disagreed with the patentee’s approach, and in the absence of any verifiable data in the application as filed, found that the effect of protein kinase inhibition was not credible. This approach places the burden of proof on the patent applicant to show that their invention is plausible, rather than on the opponent to show that there is reason to doubt the plausibility.
The questions asked of the parties during the appeal hearing make it clear that the Board was focused on the question as to whether the information presented in the application as filed was enough to establish plausibility or whether verifiable evidence is needed. The Board also appears to have concluded that plausibility must be assessed solely on the basis of the application’s disclosure, and not on data or evidence generated or submitted after filing.
In actual fact, as well as the post-filed data showing that dasatinib was active, it also showed that certain of the other compounds disclosed in the patent application were inactive. While this may have influenced the Board’s assessment as to the veracity of the applicant’s statement of activity in the application, this will not form part of the Board’s decision. However, one cannot help but speculate that this played a part in the Board’s assessment.
According to BMS, it also had the relevant data from before the filing date. Despite referring to it in one of its witness declarations, it never quantified the activity and elected not to file it in these proceedings. While this failure should not have any bearing on the plausibility of the effect described in the application as filed, one wonders whether the absence of this data influenced the decision to revoke the patent.
Finally, it may be that the Board of Appeal simply did not believe that all of the disclosed compounds (of which there were 580) exhibited activity, or sufficient activity to have any prospect of being useful. The list of compounds did appear to be a systematic progression of compounds as would be generated around a basic pharmacophore as part of a normal pharmaceutical research program. Therefore, it did appear that the claimed activity was prima facie unlikely to be shown by all the compounds. The failure to include any comparative examples which did not show activity may also have brought the claimed effect into question.
We will have to wait until the Board of Appeal issues its written decision in order to see their detailed reasoning. However, it appears inevitable that this decision will raise the threshold for plausibility.
We speculate that the big question to be answered is not whether the plausibility of a technical effect is verifiable by the skilled person, but what degree of verifiability will be required. In the present case, it seems that the technical effect was theoretically verifiable – the skilled person was taught which assays to use in order to assess the protein tyrosine kinase activity of the disclosed. Therefore, perhaps the decision will set out a requirement that the technical effect must be in some way quantified, such that it is made evident that testing has actually taken place and an effect demonstrated.
That is all very well insofar as the facts apply to the present case, but how that will translate to other pharmaceutical patents and different technologies is yet to be seen.
It does appear clear that future decisions will turn on the specific facts of the case, and the refusal by the Board to refer a question on plausibility to the Enlarged Board of Appeal confirms that there can be no universal bright-line test for plausibility.
For companies considering filing patent applications in the pharmaceutical field, it is recommended that as much verifiable evidence as possible is included in the application at the time of filing. It may be that some applicants do not wish to reveal too much information about lead candidates. In such circumstances it may be prudent to at least define a threshold level of activity, such as “compounds of the invention have an activity of greater than X in assay Y.”
For companies finding themselves in the same situation as BMS, perhaps the submission of all available data, however historical, is the best approach. You cannot fix what was filed, but you can prove that you carried out testing before filing and show what the results were. This might at least give pause for thought to the decision makers at the EPO.
From an opponent’s perspective, this decision potentially strengthens an inventive step attack, particularly where an application contains no verifiable data. For patents with a similar fact pattern, it also suggests that that the EPO should be selected as the forum for invalidity proceedings, rather than the national courts. While NCE applications are usually outside the opposition window by the time their importance is recognized, it may be that ‘defensive divisionals’ could be picked off at a later date.
Companies may also want to reconsider their filing strategies in the light of this decision. It is likely that the patent would have survived an analogous validity attack at the national level. Therefore, for commercially important technology, applicants should consider filing stand-alone or parallel national applications in Europe.
Dr. Arnie Clarke works in the fields of pharmaceutical, medicinal, industrial, food and organic chemistry, with particular expertise in small molecule pharmaceutical chemistry, secondary formulation of pharmaceutical products and industrial formulation chemistry. He specialises in contentious inter partes proceedings before the European Patent Office and has considerable experience in both oppositions and appeals, having been involved in over 150 oppositions. Arnie can be contacted on tel. no. 020 7655 8500 or email email@example.com