Benitec/CSIRO Win Major Patent Battle in US

One of the sadder stories in RNAi Therapeutics history is the many years lost in the commercial development of drugs based on DNA-directed RNAi (ddRNAi) as the space had become embroiled in litigation instead of investing in the science. In a sign that this chapter may be behind us and that ddRNAi Therapeutics can now look forward to a time where the significant medical potential of the technology can be realized, the USPTO Board of Patent Appeals and Interferences (BPAI) just reversed an earlier (2008) decision to revoke the previously issued and fundamental Graham patent that is controlled by Benitec and CSIRO. This had followed a re-exam prompted by their archrival Nucleonics Inc, now bankrupt and buried on the corporate graveyard underneath its legal bills.

Benitec shares went up 50% on the news.

The Graham patent as issued first in 2003 broadly covered the use of double-stranded DNA capable of driving the expression of double-stranded RNAs (dsRNA) for sequence-specific gene silencing in animal cells. The compositions included the expression of sense and antisense RNAs from either separate promoters or in the form of a self-complementary hairpin RNA from a shared promoter. Because the latter construction (single promoter) is the commercially more valuable claim, it will be the focus of my following discussion.

The previous re-examiner had rejected these claims as obvious in light of the famous Fire and Mello studies where the Nobel Laureates (btw, watch out for announcements next week- microRNAs this time?) applied dsRNA (not dsDNA) directly to animal cells and showed that it is in fact dsRNA and not antisense that is by far the more potent trigger for gene silencing. In a mixture of valid scientific speculation and boilerplate legal language, the Fire-Mello patent also contemplated dsRNA generated by transcription in animal cells as well as self-complementary RNAs as variations of the dsRNA RNAi trigger theme. The patent, however, does not provide examples that these methods in fact would work.

Specifically, this examiner argued that the claim in Graham covering the one promoter with the two identical gene copies in inverted orientation and separated by a ‘stuffer’ (i.e. driving the expression of hairpin RNAs) was obvious over Fire in light of antisense literature that described the use of hairpin structures at the end of antisense molecules so as to stabilize them from exonucleolytic degradation. The examiner therefore concluded that it would have been obvious to modify the Fire dsRNAs with hairpins to similarly stabilize them from degradation.

The new examiner, however, concluded that this is not so. In fact, since Fire-Mello already specifically stated that dsRNAs are naturally stable, more stable than antisense, changing this stable structure with things like hairpins would only have risked adversely affecting the stability of the dsRNA. That is, Fire-Mello in fact taught away from such modifications. Moreover, the dsRNA portion of the hairpin elements generated by the antisense stabilization technique would not have had the capacity for being silencing triggers themselves, whereas in Fire-Mello the dsRNA was the silencing trigger. Indeed, in order for the antisense to work efficiently, the target mRNA would first have to displace and disrupt the protective hairpin at the end. Thus, the hairpins in Graham and the antisense literatures served entirely different functions (Graham: efficient production of dsRNA; antisense: stabilization), and to draw such parallels was inappropriate.

In short, the USPTO found that RNAi is Not Antisense and reversed its earlier decision. As a result, when it comes to RNAi Therapeutics, the two fundamental patents are now Fire-Mello for the application of dsRNA to the target cells, and Graham for ddRNAi approaches.

Given the importance of the US market for the pharmaceutical industry, this decision has a number of implications for the RNAi IP landscape. First, of course, is that ddRNAi Therapeutics would likely require a license to Graham for their commercialization until ~2018 (priority date for Graham: 1998). Currently, there are two ddRNAi candidates in phase I clinical development, one for HIV by Benitec itself, and one for the treatment of cancer by US-based Gradalis; another one is a ddRNAi development candidate for HCV that originated with Benitec and is now with Tacere/Pfizer and appears close to the clinic. Considering the time it takes from discovery to the commercialization of new drug candidates and the so-called ‘research exception’, the demand for Graham as a gate-keeping ddRNAi therapeutic patent may therefore be somewhat limited.

Of more immediate financial benefit to Benitec could be the research and reagent market where there are a number of companies that have been selling ddRNAi vectors and transgenic RNAi mice and have yet to obtain a license from Benitec/CSIRO. Because of the complexities of the Benitec-CSIRO-Sigma relationship, we probably have to wait to hear more from the company about the anticipated financial impact here.

An even larger financial windfall would probably occur if CSIRO can get their ddRNAi patent issued for the plant field- despite Fire-Mello. It is CSIRO's dedication and attention to detail that was a major force in achieving the Herculean feat of turning around Graham.The financial windfall is due to the fact that, unlike therapeutics, ddRNAi plants are already a commercial reality and growing.

It is of note here that an interference proceeding against the Fire-Mello patent has been initiated by CSIRO and might even result in invalidating Fire-Mello altogether. In interference proceedings the aim is to determine the priority of patents (here Waterhouse vs Fire) that compete for coverage of the same subject matter, or at least subject matter that the USPTO holds to be the same. I indeed believe that there is a good chance that ddRNAi for plants will be found to have been conceived before Fire-Mello was. It is also possible that, in the end, Fire-Mello and Waterhouse will find a way to peacefully co-exist.

In the end, it is a relief that the USPTO did not lose sight of major themes in RNAi science and did not get lost in technical minutiae. It therefore also bodes well for should the time come that the Crooke antisense patents are moved to the frontlines of the patent battles. But wouldn’t it be great to wait next time until real drugs have been developed before fighting over the spoils?

PS: It seems like there have been major developments on another prominent RNAi patent front, the Tuschl Litigation. As reported on the RNAi Litigation blog, it appears as if UMass is becoming increasingly isolated, and any attempt to rescue the therapeutic value of Tuschl-I is getting less likely by the day.