RNAi Therapeutics at ASCO 2011- A Preview

This year’s ASCO meeting will feature two important RNAi Therapeutics clinical studies: Silence Therapeutics’ Atu027 for advanced solid cancers, and Alnylam’s ALN-VSP02 for cancers with liver involvement. The data presentations kick off the most important period of clinical newsflow in the history of RNAi Therapeutics. Over the next couple of months, first clinical data from the best crop yet of systemic RNAi Therapeutics candidates will be announced for indications ranging from cancer, TTR amyloidosis, and hypercholesterolemia. The importance of the data is amplified by the fact that these candidates employ delivery systems for which additional candidates are already in the clinical queue.

ASCO will be first about the safety of ALN-VSP02 and Atu027 and their underlying delivery systems. In this regard, the phase I studies may be considered successes if the reported or anticipated maximally tolerated dose levels are well into the range where based on the preclinical data anti-tumor efficacy would be expected. Second, it will be important to see evidence that these drugs act via their predicted mechanisms of action. Needless to say, data hinting at preliminary efficacy would be a definite bonus.

Atu027: Will dose escalation continue to pierce through the predicted efficacious dose?

Atu027 targets PKN3, a gene scientists at Silence Therapeutics first discovered to be involved in cancer-related processes downstream of the more widely recognized PI3K pathway (more on the background of Atu027 here). Although the precise mechanism remains somewhat unclear and could be relatively complex, as so often is the case in biology, various lines of evidence point towards an important role of PKN3 in the metastatic spread of tumor cells.

The phase I study of Atu027 to be presented at ASCO is an open-label, multi-dose dose escalation trial in patients with advanced solid tumors. Enrolment started at the end of 2009 at the low dose of 0.001mg/kg. This low starting dose was a consequence of this being the first human experience with the Atuplex delivery system and the known potential of cationic lipoplexes to stimulate various immune responses.

The trial experience, however, suggests that this is not an insurmountable challenge, and in fact, the safety profile of Atu027 so far has been very encouraging. As of the last update (Silence Therapeutics financial update conference call in April 2011), 24 patients had been enrolled in 8 dose cohorts. The only noted safety issue was a transient activation of the alternative complement pathway, something which was not surprising for this positively charged formulation, but which the company characterized as not being clinically meaningful. Accordingly, after more than 170 dose administrations, no dose limiting toxicities had been reported. The company also remarked that, unlike competing technologies, Atu027 does not make use of pre-medications in order to hedge against potential, or expected, immune stimulations, obviously pointing the finger at ALN-VSP02 (see discussion below). Indeed, in a foundational paper on Atu027 (Aleku et al. 2008 in Cancer Research), the company demonstrated pretty convincingly the absence of cytokine responses.

In terms of efficacy, it is relevant that the study has now entered dose ranges where based on the preclinical pharmacology in non-human primates and rodents, knockdown and anti-cancer efficacies can be expected. Considering the apparently benign safety profile, this in itself is highly promising and I look forward to seeing the investigators continue to dose escalate and wrap up the study this fall.

There have been 6 patients that remained stable 3 months after their enrolment in the trial. There is also one anecdotal report of tumor regression:

As can be seen on the CT on the left, a ~2 centimeter metastatic lesion in the lung of a cohort 6 patient disappeared one week after receiving 8 infusions of Atu027 (left/right panel: before/after). It is worth bearing in mind though that this trial design may not best capture the utility of Atu027 as its primary benefit may be not only in inhibiting or reversing primary tumor masses, but in preventing their metastatic spread.

Overall, this study appears to be on a very promising trajectory and I look forward to a more thorough discussion of the updated data at ASCO.

ALN-VSP02: Clearer picture of the safety profile and mechanistic data

Unlike Atu027, the SNALP-formulated ALN-VSP02 targets tumor cells, taking advantage of the Enhanced Permeability and Retention (EPR) effect of nanoparticle delivery to solid tumors. It is also of particular interest that this drug candidate targets two validated cancer genes simultaneously: kinesin-spindle protein (KSP) to inhibit cell division and VEGF for anti-angiogenesis. In principle, I like this dual targeting quite a bit, also as it blazes the regulatory trail for other multi-targeted siRNA therapeutics not only in cancer indications (more on the background of ALN-VSP02 here).

The multi-dose phase I study of ALN-VSP02 has enrolled over 40 patients and escalated up to the solid dose of 1.5mg/kg. Three acute infusion reactions were noted, but these were readily managed by adjusting the infusion rate and did not preclude future dosing. There were, however, two dose-limiting toxicities. One was a death, the other a grade 3 thrombocytopenia. The thrombocytopenia occurred at the relatively high dose of 1.25mg/kg so this should not be overly concerning. The other was a death at the 0.7mg/kg dose. As I had discussed in a previous blog entry, it is difficult to tell what precipitated this death. However, as it was possible that this was related to the drug and the extent of the metastatic tumor burden in the liver, subsequent enrollment criteria were adjusted to limit the extent of liver cancer.

The trial already shed light on one important aspect of the SNALP delivery platform (developed by Tekmira): the feasibility of repeat-administration. Before this study, it was unknown whether repeat-administration would affect pharmacology and safety through adaptive immune responses. This for example was a particular issue for first-generation monoclonal antibodies, and, to some degree, still is. It is therefore a great relief that the pharmacokinetics of ALN-VSP02 are apparently unchanged following repeat-administration. It is little things like this, or the ability to manufacture at large quantities, that will decide whether a delivery platform will ultimately be successful.

In terms of efficacy, the company reported at last year’s ASCO that the blood perfusions of liver tumors were significantly reduced in a majority of patients treated with ALN-VSP02. This would be consistent with an anti-angiogenic effect as a result of VEGF knockdown. Given, however, that this was not a dose-proportional effect, it is probably premature to conclude that this was due to on-target activity.

This is one of the reasons why I don’t really like the choice of VEGF as one of the target genes. Cytokines are known to have the potential to affect angiogenesis, sometimes inhibiting, sometimes promoting it. The origin of ALN-VSP02 fell in a time when Alnylam’s target selection was less developed than it is today. It also fell in a time when immune stimulation was blissfully ignored by many if not most players in the field and tensions with Protiva/Tekmira were quite high. Tekmira had long recognized that almost as important as delivering the siRNA to the right place is the need to carefully characterize and mitigating liposomal nucleic acid-triggered cytokine responses, and are to be considered experts in this area.

In the absence of a peer-reviewed publication on ALN-VSP02, I have looked through the related patent literature and conference and company presentations, but could not find an immunological evaluation of ALN-VSP02 as it has become industry-standard today. And as Silence Therapeutics has been hinting at, it is also notable that the study protocol of ALN-VSP02 involves concurrent treatment with steroids, H1 and H2 blockers, and acetaminophen. To be clear, such drugs of often used to manage potential infusion reactions which are observed with many systemically administered chemotherapeutics and monoclonal antibodies, but it makes me wonder a bit.

I’m just putting this out for everybody to form their own opinions, and maybe for Alnylam to address some of these issues in future presentations. I would also stress that in SNALP delivery, it is not about whether cytokine responses are induced or not, it is about characterizing and mitigating/abolishing them.

It is therefore all the more important for Alnylam to provide more evidence that the main mechanism of action of ALN-VSP02 is via RNAi target cleavage. In this regard, important data were presented earlier this year when Alnylam showed in a carefully controlled experiment that ALN-VSP02 triggered bona fide RNAi cleavage in liver cells obtained from biopsies. Moreover, in similar biopsy analyses, the company showed drug levels in the liver that are consistent with RNAi knockdown activities based on the extensive preclinical experience with SNALP delivery.

Overall, ALN-VSP02 looks like a decent candidate and it is clear already that RNAi is operative. The trial organizers have also done extremely well in escalating the study up to 1.5mg/kg and milking the study for data, all the while protecting the safety of trial participants. I therefore look forward to learning more about the mechanism of ALN-VSP02 in Man such as cytological analyses demonstrating spindle abnormalities as would be expected from a KSP knockdown. Of course, any type of tumor response and regression would be a major plus.