Big Data, products & processes: being a German patentee in the era of the Rezeptortyrosinkinase decisions

The IPKat trying to put his paw on which of the data
may or may not be infringing....

Here is a pretty uncontroversial prediction - we are going to see more infringement cases concerning products obtained directly from a patented process where the product ends up being data, for example from test results that are used as the basis for diagnosis or treatment.  But with data entering a jurisdiction from a processes occurring abroad, what is a poor patentee to do (especially when fighting with the loss of identity test in Pioneer v Warner)?

Kat friend and cross-border colleague Jérôme Kommer in the Mannheim office of Quinn Emanuel reports on the current position in Germany following recent Supreme Court decisions.  Over to Jérôme:

"No doubt about it, we are living in the era of Big Data solutions used for generating and discovering information that has been previously unavailable or prohibitively difficult to obtain.  With the merging of data analysis and other technologies, such as gene sequencing and diagnostics, where does that leave litigants, especially in cases in which the information as such may have its own, potentially considerable, value?  

Last year, the German Supreme Court had the chance to address these questions in two decisions involving the same patent, EP 0 959 132 B1.  The '132 patent covered a nucleic acid molecule encoding a receptor protein kinase (“fms-like tyrosine kinase 3” or “FLT3 gene”, in particular) and a method for detecting the claimed nucleic acid molecule. The detection can be used to diagnose leukemia, for example.

In the first decision (Rezeptortyrosinkinase I of January 19, 2016), the Supreme Court found the subject matter of the claims directed at the nucleic acid molecule patentable subject matter.  This was despite the fact that the claims did not contain any language indicating that the nucleic acid molecule needed to be isolated or created by a particular technical process. In the Supreme Court’s opinion it was sufficient that product claims necessarily (and inherently) require that the product is provided by a technical process (para. 51). In addition, it was not necessary to identify an “inventive concept” to distinguish patentable subject matter from unpatentable discoveries (para. 50, distinguishing German and European patent law from the US Supreme Court’s Mayo v. Prometheus, 566 U.S. __ (2012)). Because it referred to these product claims, the method was also patentable (para. 53).

The importance of the second decision (Rezeptortyrosinkinase II of September 27, 2016) in the parallel infringement case goes far beyond the already important issues of patentability surrounding gene sequencing and its use to diagnose diseases of the first decision.  Because the defendants practiced the protected method outside of Germany and just sent the results of the diagnosis back to their German customers, the defendants could only infringe the German part of EP 0 959 132 B1 if these results constituted a product directly produced by the protected process (“unmittelbares Verfahrenserzeugnis”, § 9 No. 3 German Patent Act).

The Supreme Court denied infringement and held that:

(1) the results, as a sequence of data, could only be a product directly produced by a protected process if the sequence of data has product-like technical characteristics provided by the process such that it could be subject to a product claim, and

(2) the presentation of the results and any findings based on the process are not a product according to § 9 No. 3 German Patent Act because it is just presentation of information that is excluded from patentable subject matter, § 1 (3) no. 4 German Patent Act, Art. 52 (2)(d) EPC.

The reference by the Supreme Court to and affirmation of its earlier decision in MPEG-2-Videosignalcodierung (a case about video images encoded by a patented encoding method that were stored on DVDs) shows that the new decision is technologically neutral.  The decision not only applies to test results obtained through genetic tests, but equally to other processes producing data and, thus, information.

The Supreme Court successfully walks the line between providing protection for a technical teaching protected by the respective process claim and not “excluding the exclusion” from patentability for the untechnical mere presentation of information. For this purpose, the Supreme Court distinguishes the data structure from the information contained within the data (para. 21 of the decision). The data structure which results from the patented process can be patentable, just like a method about how to display moving images on a TV set is patentable. On the other hand, the information itself, i.e. the specific moving image, e.g. a video of a horse race, cannot be patented because it is not technical (cf. para. 22, 24 of the decision).

Applied to the nucleic acid molecule detection method, this meant that the only “product” of the method was a biochemical result about the presence or absence of a specific mutation indicating leukemia or not (para. 19 of the decision). Only that result - presence of leukemia or not - together with some supporting information and not the original data sequence were sent to the German customers. Because there was no specific (and even a less technical) way of presenting this information or any other characteristics obtained from the information obtained form the patented method, these findings did not amount to a use of the patented invention (para. 24 of the decision). Therefore, the Supreme Court denied protection under § 9 No. 3 German Patent Act.

The two Rezeptortyrosinkinase decisions clarify the answers to the questions mentioned at the beginning, contain specific guidance for the protection of test results based on gene sequencing, and strike a balance between the accessibility of patent protection for such methods while excluding resulting mere information from patent protection.

Beyond these two decisions, it will be interesting to see the further development of the Supreme Court’s general acknowledgement of the patentability of data sequences. That data sequences need to be able to be used an arbitrary number of times just like physical products and can constitute commercial goods, which is generally the case if they can be reproduced on physical data carriers and/or transmitted via the Internet (MPEG-2-Videosignalcodierung, para. 22, 23), is not much of a hurdle. The more interesting and decisive test for the scope of product-by-process claims and patentability of data generating methods relates to the technical nature of the respective claim’s subject matter."

Is the message to patentees to patent a data sequence or structure containing the abbreviated results? Would that have saved the patentee? Jérôme responds:

"If there was a data sequence with sufficient technical characteristics that could have been made subject of a claim (e.g. with a data structure or “syntax” that at least contributes to the solution of a technical problem), that could have been a way to (also) protect the commercially valuable information contained within the data sequence. Of course, the data sequence as the protected product would have to be novel and inventive, too. However, it is unlikely that that would have been possible in the case decided by the Supreme Court. And even though the information as such might have its own (considerable) commercial value, patent protection should be limited to the novel and inventive technical teaching provided by the patentee. Here, this was the method of identifying the claimed nucleic acid molecule, which could have been protected outside Germany if the patentee had applied for patent protection in the country in which the method was practiced."

The AmeriKat also understands that there was a case heard last week before the Supreme Court in Germany about the liability for patent infringement in Germany if the infringer "only" delivers infringing products to customers outside Germany which then might end up in Germany.  The case no is X ZR 120/15.  Watch the IPKat for future updates.