ALN-VSP02 Interim Safety and Early Signs of Activity Supporting Continuation of Trial

[This entry has been updated on May 21, 2010]

At precisely 6pm tonight, ASCO released the much-awaited abstracts for presentations at the upcoming cancer meeting. This included one on the interim analysis of Alnylam’s ALN-VSP02for the treatment of liver cancers (my review with Tobias Wolfram on VSP02) targeting both VEGF (antiangiogenic) and KSP (anti-proliferative) and delivered by first-generation SNALP technology*. Importantly, the drug showed an encouraging safety profile with no cases of flu-like hypersensitivity reactions up to the highest dose tested so far (0.7mg/kg, dose-escalation ongoing) and signs of hepatotoxicity up to 0.4mg/kg, the two most important expected potential toxicities for SNALP delivery-based drugs.

There was one death possibly related to the study drug in a patient with a pancreatic neuroendocrine tumor (most others had colorectal cancers) at the 0.7mg/kg dose level. This patient died of hepatic failure after receiving the 2^nd of the 4 bi-weekly 15-minute infusions. While liver patients do often die from hepatic failure and patients in the trial probably have a relatively short 1-2 year life expectancy which means that deaths are to be expected in such a trial just by chance, because of the temporal association of the treatment with the death, a contribution of the drug cannot be excluded. A general hepatotoxicity of the drug or delivery technology, however, is unlikely because, unlike flu-like reactions to liposomal siRNAs, hepatotoxicities should be more uniform, at least in 'average' persons, but were otherwise absent in the 0.4mg/kg cohort as well as in the two other patients dosed at 0.7mg/kg. Similarly, except for a minor grade 2 infusion reaction that responded to slowing of the infusion, there were otherwise no significant adverse events, quite remarkable for such a cancer trial, which supports that the death was an isolated case of which the exact cause remains to be fully determined.

Interestingly, in the deceased patient there was extensive tumor necrosis following drug treatment and this was correlated with decreased blood flow as measured by DCE-MRI, consistent with successful VEGF inhibition. Although it cannot be excluded at this point, it is probably too early to say whether the death may have even had to do with too rapid tumor necrosis due to VEGF and/or KSP inhibition by affecting already impaired liver function in liver cancer patients. Altogether, blood flow was measured in 8 of the 12 patients treated as of December 2009 with 80% of the tumors showing a remarkable >40% decline [in blood flow].

Eyes are now on the more detailed data presentation at the ASCO meeting in June that will include further safety, tolerability and also pharmacodynamic data, possibly at even higher dose levels**. More information on the exact distribution of the declines in blood flow and the neuroendocrine patient history will be of particular interest.

* Second generation SNALP formulation have potencies that are about 10-50 increased compared to the one used in the present trial that had a predicted 50% knockdown potency at around 1mg/kg. This means that at therapeutically relevant doses for future SNALP-based drugs of around 0.05-0.2mg/kg, no significant toxicities have been observed so far in both Tekmira's SNALP-ApoB and Alnylam's ALN-VSP02 trials.

Putting the ALN-VSP02 Adverse Event into Context (update May 21, 2010)

It has come to my attention that there is considerable angst, especially among investors, about the one death in the trial. While every such case is unfortunate and has to be studied in detail whether it is linked to treatment with the investigational drug, let me emphasize that the reason why liver metastases are specifically treated for most cancers, including in this trial, is because they turn out to be rate-limiting for many of these late-stage patients. As such, more deaths are to be expected in this trial, also and especially due to liver failure. One caveat, however, is that the patient died after the 2nd infusion, that is between 2-4 weeks after dosing had been initiated and one of the enrollment criteria for the trial was a life expectancy of >12 weeks, so a contribution of the drug is likely.

A good indication about the seriousness of these adverse events as it relates to drug safety and risk:benefit is how regulators react to such reports that by the way have to be made in a timely manner, meaning that if there had been more such reports they must not have been deemed sufficient to stop the trial. Related to this, it will be important to learn whether the regulators allowed further dose escalation or whether the patients that have been recruited since the December 2009 abstract deadline were treated largely at the 0.4 and 0.7mg/kg dosages to further study the drug safety around these dose levels, although the company did say in their press release that the maximum tolerated dose has not been 'reached' yet [update: the company has since confirmed that dose escalation has continued since December 2009, meaning that at the very least there should now be data for the 1.0mg/kg cohort if not higher which reflects favorably on the safety profile thus far].

It is surprising that Alnylam did not comment generally much about the detailed data in the press release about the abstract. It is possible that they consider doing so at this point as being of little value since much more comprehensive and informative data will be presented soon at the ASCO.

In the meantime, unlike drugs developed maybe for restless legs syndrome or ED, deaths are a frequent occurrence of cancer drug development and investors have to live with that without assuming the worst.

ALN-VSP02 with Potential to Become Valuable Targeted Component in Liver Cancer Therapy

In the final part of our 3-part series looking at the 3 most advanced RNAi Therapeutics candidates for cancer (part 1: CALAA-01; part 2: Atu-027) Tobias Wolfram and I have been analyzing ALN-VSP02, Alnylam’s candidate for the treatment of cancers with liver involvement that has entered the clinical stage of development in the first half of last year. ALN-VSP02 is a 2-pronged strategy to push back liver cancer comprising of 2 siRNAs packaged in Tekmira’s SNALP delivery formulation, one directed against the well validated vascular endothelial growth factor (VEGF) to choke off the nutrient and oxygen supply to the liver and the other against kinesin spindle protein (KSP) to disrupt cell division and induce apoptosis. Overall, based on the pre-clinical data and scientific rationale of the approach, we consider VSP02 to be a solid clinical candidate with potential to become an important component in the fight against a disease for which new options, especially molecularly-targeted ones are desperately lacking.

Liver cancer, both primary and secondary, is one of the most underserved cancers for which new therapeutic approaches are urgently needed. According to the 2009 issue of ‘Cancer Statistics’ by Jemal and colleagues, almost as many people die of primary liver cancer as are diagnosed. Primary liver cancer (aka hepatocellular carcinoma/HCC) has only recently risen to prominence in Western societies with ~22,000 newly diagnosed cases in the US alone in 2009, partly the result of an increased Asian immigrant population and the hepatitis C wave starting to take its toll. It is an even much larger problem worldwide with about 500,000 annual new cases particularly in the rapidly growing countries in Asia where hepatitis B infection is so prevalent. In addition to primary liver cancer, it is metastatic liver cancer that is the cause of much mortality arising from cancers of non-liver origins. Often more aggressive than primary HCC, such cases account for about 50,000 of newly diagnosed cases in the US with colorectal metastatic to the liver accounting for about 80% of those.

If you are lucky, your liver cancer is a candidate for surgical resection which will extend your life expectancy significantly and in a few cases is even curative. Unfortunately, due to the disseminated nature of liver cancer and the poor health of the liver at the time of diagnosis, surgical resection is not possible for the majority of cases (only 10-20%). One objective for the development of new drugs has therefore been to shrink the cancer sufficiently that patients become eligible again for resection. Despite some success with (systemic) chemotherapy, these often suffer from dose-limiting toxicities and are not curative. Liver transplants and relatively crude methods involving burning up the cancer tissue with heat or radioactivity are frequently used alternatives, but as you can guess, are either rather desperate attempts at fighting liver cancer or not practical for large patient populations.

There are, however, also reasons to be hopeful. Systemically administered sorafenib (aka Nexavar; first approved for kidney cancer) for example is the first ‘targeted’ therapy approved for primary liver cancer, a small molecule 'targeting multiple’ kinases with varied functions including angiogenesis. Illustrative of the high unmet need in HCC, a pivotal trial with this drug was halted pre-maturely to make the drug rapidly available to patients after it has shown an increase in median overall survival from 34.4 weeks on placebo to 46.3 weeks and median time to progression from 2.8 to 5.5 months. Unfortunately, most other small molecules and chemotherapeutics have not met with similar success largely because of drug resistance and systemic toxicities.

The avoidance of systemic toxicity is also the reason why a recent development in liver cancer has been quite exciting and that is poised , together with surgical resection, to set a newstandard-of-care for many types of liver cancers: a regional therapy for chemotherapeutics (and possibly other agents) that is repeatable and works by isolating the hepatic circulation and thus allows for bathing exclusively the liver in cytotoxic agents. The developer of this drug-device combination, Delcath Systems (ticker: DCTH), has just shown very promising top-line data in a phase III study for melanoma metastatic to the liver (another one of those cases where the liver metastasis is the major cause of mortality) extending hepatic progression free survival from 70 days with ‘best alternative care’ to 217 days with the company’s percutaneous hepatic perfusion system, aka PHP(detailed data to be presented at ASCO in early June; disclosure: DH owns DCTH shares). Other studies using this system for primary liver cancer and secondary colorectal and neuroendocrine are currently in phases II and III of clinical development. Possibly a lesson for Alnylam for future studies is that Delcath plans to conduct much of its late-stage clinical development for primary liver cancer in Asia, both for patient access and eventual market, while the trials for metastatic liver cancer largely take place in the US. Nevertheless, even if PHP will become part of a new standard-of-care for liver cancers, the problem is far from solved with hepatic disease in many cases eventually recurring and non-hepatic sites becoming rate-limiting. For cancer, monotherapy is seldom the answer, and for liver cancer this could mean a combination of surgical resection, regional high-dose chemo, plus one or two targeted therapeutics such as ALN-VSP02.

In this context, an RNAi Therapeutic should aim to either reduce/remove residual tumors after surgery and/or chemo, (re-)sensitize tumors to chemo, or contribute entirely new mechanisms of actions to the treatment of liver cancer. In addition, because a SNALP particle such as in ALN-VSP02 has selective delivery (main sites: liver, and other sites of solid cancers including lymph nodes) and the siRNAs target genes specific for cancer, it may be a systemic therapy with potential to also address sites of extra-hepatic disease.

ALN-VSP02 has the potential both to sensitize liver cancers to chemotherapy and to hold the cancer in check by itself by starving it and inhibit cell proliferation. Much of the discussion that I had with Tobias was centered not so much on delivery, but on whether the VEGF siRNA component was the best choice. There is little doubt from the broad success with Roche’s VEGF blocking monoclonal antibody Avastin in a variety of solid cancers that the angiogenic (blood vessel growth) factor VEGF should be a good target also for the typically highly vascularized liver cancers. Our question, however, was that since there is already Avastin, maybe a gene more uniquely suited for RNAi Therapeutics may have been preferable. On the other hand, as the first VEGF targeting RNAi Therapeutic, ALN-VSP02 is not simply a me-too drug because of a different mechanism of action: preventing VEGF from being made locally instead of blocking it once made and then also systemically (which causes additional safety issues). This could lead to unanticipated treatment benefits, but by the same token of course also potentially unanticipated failures. Questions that remain to be answered are for example how a tumor will respond to a maybe 50-70% overall knockdown with intratumor variations in silencing efficiencies as can be expected for SNALP delivery to solid tumors. And even if it ‘only’ had comparable efficacy to Avastin in terms of inhibiting angiogenesis, the fact that VSP02 addresses two targets at once means that it has the potential to substitute for Avastin as there are only that many drugs that a given patient can take.

The siRNA targeting kinesin spindle protein (KSP) is the less controversial component of VSP02, although it also falls in the category of a ‘druggable’ target under traditional definitions. Accordingly, almost all pharmaceutical companies to speak of have created their own, often me-too KSP-targeting small molecules, a number of which are in phase I and II clinical development for a variety of cancers. KSP is an attractive target because its tubulin-organizing function is thought to be specific to mitosis (the last stage of cell division) and its inhibition is not expected to cause side-effects typically associated with commonly used anti-cancer drugs that bind tubulin directly (e.g. neurological and hematological side-effects). Compared to these small molecules, however, ALN-VSP02 should have the added benefit of enhanced specificity and potency also because of its more selective delivery to solid cancers compared to small molecules. Moreover, many of the small molecules have IC50s in the high nM and low microM range, significantly higher than ALN-VSP02.

ALN-VSP02 has been very well validated in pre-clinical studies demonstrating expected pharmacological effects in mouse models of primary and metastatic liver cancer: compromised spindle bodies (‘monoasters’) from KSP knockdown and a decrease in microvessel density and vascular leakage from VEGF inhibition. In addition, Tobias was immediately struck by the maturity of the SNALP delivery system underlying VSP02. Rodent studies, of course, are necessary to evaluate the pre-clinical activity of an anti-cancer drug, but SNALP distinguishes itself in that its safety and efficacy has been routinely confirmed in non-human primates, dating back to 2006.

Liver cancer is a peculiar application for SNALP delivery technology for sure. On the one hand ‘liver’ suggests that first-generation, short circulating SNALPs may be appropriate. On the other hand, liver cancer tissue is different from the normal liver parenchyma in that it is quite a bit more heterogeneous, poorer in cell content and higher in extracellular matrix, and is thought to be primarily supplied with nutrients and oxygen by the hepatic artery rather than the portal vein as is the case for normal liver. It is possible for this reason that the PEG-lipid anchor in VSP02 which largely determines circulation times has a long carbon chain (18C). Such a long-chain lipid anchor has the added benefit in that it may also be able to address co-existing cancer outside the liver. In fact, research published last year by Tekmira showed that in the case of liver cancer short and long chains have comparable efficacy, whereas for cancer outside the liver long chains are preferable. The ability of VSP02 to also address non-liver sites of cancers was then also demonstrated by Alnylam late last year [erratum: while we originally believed VSP02 to comprise a C18 long-chain PEG-lipid anchor based on the schematics in Alnylam's poster presentations, the pharmacologic data presented at ASCO 2010 and the actual formula- PEG2000-cDMA- strongly suggest that it is in fact a short-circulating C14 myristyl anchor].

Tobias noted, however, somewhat critically that ALN-VSP02 is not an actively targeted therapeutic to which my response was that at least it is passively/pharmacokinetically targeted, and that Tekmira and Alnylam are working hard on next-generation formulations with targeting ligands that promise increased targeting selectivity and lower doses. This, however, is at the expense of more complex formulation methods and will take time to develop. We then quickly got into a discussion about the importance of developing such first-generation drug candidates in general, and we soon agreed that their value also lies in providing the foundation for more potent and specific follow-ons (not only targeted SNALPs, but also more potent lipids with a larger therapeutic index), at the end of which there may be cancer treatment strategies without the need for chemotherapeutics altogether.

The ongoing phase I trial is an open-label, non-randomized study aiming to enroll about 55 patients with primary and secondary liver cancers. First results on the pharmacology and biomarkers are to be presented at the upcoming ASCO meeting. This presentation should provide important insights into the functional knockdown of VEGF and KSP, and the future development path of this drug candidate. In addition to differentiating between primary and secondary and geographic differentiation of future trials (-->Takeda for Asia?), additional biomarkers (e.g. based on mutation status of a range of cancer-related genes or more dynamic biomarkers such as microRNAs) may help to further dissect the patient populations into those that are expected to respond best to VSP02 and select the most promising companion drugs for phase II studies. In the end, while the rationale and pre-clinical results are sound, only clinical experience will tell to what extent an innovative drug candidate such as VSP02 will be able set in motion the complex chain of events leading to the reduction or even destruction of liver cancer.

Opko Health Announces Initiation of First Phase III Trial of an RNAi Therapeutic

Less than a week after the signing of two significant alliances between RNAi Therapeutics companies with Big Pharma, it is today up to Opko Health to announce the initiation of the first phase III clinical trial for an RNAi Therapeutic. Opko Health is an ophthalmic focused company that was recently formed through a string of mergers and acquisitions with one of the companies involved being Acuity Pharmaceuticals. It is from Acuity that Opko inherited two advanced clinical RNAi Therapeutics programs, one of which is the subject of today’s announcement.

In the proposed COBALT study, Cand5 (bevasiranib), an unmodified siRNA against VEGF, will be given once every 8 or 12 weeks in patients with wet age-related degeneration (AMD). The goal is to assess its safety and, more importantly, whether it has equivalent efficacy compared to a the currently leading wet AMD drug Lucentis, which is another VEGF inhibitor that is given once every 4 weeks by needle injection.

Although the sweet spot for RNAi Therapeutics are targets that are otherwise undruggable by small molecules and monoclonal antibodies, Opko is one of a handful of companies targeting the VEGF pathway for AMD and diabetic retinopathy (see “RNAi and the Eye” post on May 1, 2007). This is in spite of the fact that other widely prescribed drugs, namely the monoclonal antibody Lucentis and the RNA aptamer by OSI Pharmaceuticals already serve this market. While this may reduce development risk and function as a proof-of-concept, RNAi Therapeutics for these applications will have to compete directly with such therapies in terms of safety and tolerability, potency, and duration of efficacy.

The reason why Opko wants to challenge Lucentis on duration is because each needle injection carries a risk of damaging the eye and causing discomfort to the mostly elderly patients. This becomes a particularly pressing issue for a repeat-administered therapy such as for wet AMD. Therefore, being able to reduce the frequency of injections by half or even more without a loss in efficacy would make an RNAi Therapeutics a very attractive treatment option. Indeed, pre-clinical studies published last year on the silencing of liver-expressed ApoB100 by systemic administration (Zimmermann et al. (2006) Nature 441: 111-4) support the notion that RNAi Therapeutics may have comparable or even better pharmacokinetics compared to what is usually observed for therapies such as monoclonal antibodies

While a positive outcome would certainly help the field of RNAi Therapeutics, there is cause to be skeptical. In particular, bevasiranib is an unmodified siRNA that is given without a particular performance enhancing formulation. This may result in suboptimal gene silencing due to RNA instability issues and inferior cell delivery and ultimately exhibit poor pharmacokinetics. Indeed, results from the C.A.R.E. phase II studies in 129 wet AMD patients were mixed and did not show statistically significant evidence for improvement of acuity. Opko clearly sees the need for optimizing RNAi delivery and have two years ago formed an alliance with the RNAi nano-delivery company Intradig to develop topical and other formulations for Cand5. Whether this will directly impact the current studies is unclear.

I am therefore more optimistic about the approach taken by Merck (formerly Sirna Therapeutics) and their partner Allergan to develop a slow-release formula of a modified siRNAs against the VEGF-receptor that when combined may significantly reduce the need for frequent needle injections. Phase II studies for that trial have started earlier this year.