Ebola Emergency Tests Oligo Manufacturing Readiness

With officials scrambling to put in place measures to contain the further spread of Ebola, it is highly likely that RNAi Therapeutic TKM-EBOLA will be on their shopping list.  This, however, requires the ability to manufacture the oligonucleotide-based medicine in quantities sufficient to treat at least 10 thousand or so either infected or potentially infected persons.  In the absence of commercial Oligonucleotide Therapeutics success stories this could prove to be a challenge, although in this case, I believe it's doable.

Manufacturing, an advantage of TKM-EBOLA over PMOs and antibodies

When the US Department of Defense selected TKM-EBOLA as its preferred Ebola development project a few years ago, manufacturing, and not just efficacy and safety should have been part of the equation.  The reason is that e.g. currently about 0.3mg/kg*70kg/day*7 days= i.e. approx. 140mg of TKM-EBOLA RNA oligonucleotide is required per treatment course.  Since its simple chemistry (a couple of spiked-in standardd 2’-O-methyls in an otherwise unmodified RNA) makes it one of the cheapest RNA oligonucleotides conceivable, let’s ballpark its manufacturing cost at $300 per treatment course at some of the largest possible manufacturing scales possible today (kilograms).  Note that the cost of the lipids in TKM-EBOLA is negligible compared to the oligo component.

1.4kg oligo and $3 million for 10,000 treatment courses.

Taking into account that these are the pure manufacturing costs when protocols have been established, and other monies will have to be spent when starting from sequence design---let’s say conservatively $5 million. I believe given the gravity of the current situation, this is a number we can live with (1/200 of investment in the response).

If you do the same Gedankenspiel for Sarepta’s Ebola therapeutic morpholino antisense candidate, you have to multiply the $3M number first by a factor of 30 for the much larger amount of oligonucleotide required and then by another factor of 6 or so for the greatly increased costs of making morpholinos over standard RNAà $500M+.  A non-starter not just for the increased costs, but also because the manufacturing capacities for that amount of oligonucleotides are not readily available today, let alone for the morpholino chemistry where Sarepta has experienced significant delays in obtaining sufficient oligo supplies for a relatively small (~100 patients on drug at 30mg/kg/week for 48 weeks) phase III program in an orphan diseases indication (DMD).

When we move to Tekmira’s antibody competition, in particular ZMapp, the situation is not all that different from Sarepta’s with the slight advantage that ZMapp could somewhat tap into the established know-how of monoclonal antibody production.  But in the end, we’d be talking about years of process development and scale-up compared to a few months TKM-EBOLA.

Wake-up call for oligonucleotide manufacturing

Although TKM-EBOLA should be within current oligonucleotide manufacturing capacities, the Ebola situation should get oligonucleotide manufacturers wondering whether they are ready for the upcoming surge in Oligonucleotide Therapeutics approvals and sales.  Following years of disappointment about the lack of big manufacturing requests, CMOs have been loath to build the plants that can churn out hundreds of kilos or even tons of oligonucleotides.

Consequently, I attribute the decisions of first ISIS and Sanofi/Genzyme, then Alnylam (for GalNAc conjugates), and more recently also Sarepta, to shift oligonucleotide manufacturing in-house, to this lack of outside manufacturing capacity.  While having manufacturing in-house may sound attractive for a number of reasons, spreading manufacturing risk across multiple vendors is an accepted risk reduction strategy in this and other industries. 

I am therefore hopeful that the ongoing Ebola outbreak will end up increasing oligonucleotide therapeutics manufacturing capacities to rule out a situation that manufacturing constraints could limit commercialization of agents like ISIS-ApoCIIIRx or the HBV agents with potentially very large patient populations. 

Tekmira Expands Biodefense Efforts to Cover Both Ebola and Marburg Virus

Following the successes with its Ebola biodefense program, Tekmira has started to present increasingly promising data for treating a related filovirus, the Marburg virus.  With 100% survival rates in non-human primate models and new RNAi triggers that should cover a broad spectrum of strains and possibly related viruses, the company is well positioned to also take away the Marburg indication from competitor Sarepta which in turn has become preoccupied with its exon-skipping drug candidate for DMD.  

For the Ebola program alone, the $100M market cap company estimates that the successful development of a drug under the so called Animal Rule could result in stockpiling orders from the US government worth about $100M annually.

From guinea pigs...

In a publication last month, Tekmira and their collaborators from the UTMB in Galveston, Texas, reported the successful treatment of guinea pigs infected with a number of different strains of the Marburg virus.  Because the development of treatments for rapidly mutating viruses and viruses with a multitude of divergent strains is hampered by sequence diversity, a broader strain coverage was achieved, like in Tekmira’s Ebola approach or in Arrowhead’s chronic HepB strategy (ARC520), through the concurrent use of two siRNAs in a single formulation.

Given that rodents have limited predictiveness for anti-filoviral efficacy in Man and given that the formulation was apparently a 1^st generation D-LinDMA-based SNALP, I was not all that excited about the publication due to the apparent early stage of development.

...to non-human primates

In a positive surprise therefore, the Chief Scientific Officer of the company, Ian MacLachlan, presented gold standard non-human primate data of SNALP RNAi Therapeutics for Marburg virus at the ongoing OligoDIA regulator-industry conference.  Accordingly, Tekmira’s newer LNP formulations (‘SNALP-G’) were shown to fully protect monkeys from death due to Marburg infection when given at 0.5mg/kg (=the magic safety threshold for SNALP).

As an important comparison, Sarepta last year reported ‘83% to 100%’ protection rates in comparable models when treatment was initiated up to 96 hours after infections with its newer PMOplus morpholino chemistry.   It is therefore of interest to test the impact of further delaying treatment with SNALPs.

As I have been following Tekmira’s biodefense program over the years, one development I noticed, especially in the era of budget cuts, was the apparent push by the Department of Defense for biomergency treatments that can address not just multiple strains of a virus, but multiple viruses all-in-one.  Intriguingly, one of the bullet points in Tekmira’s slide presentation on goals for the Marburg program (slide 36) talks about the ‘Design broad spectrum siRNA to target multiple MARV and Ebola’.  Does this mean that the ultimate prize here are stockpiling contracts for a single drug addressing both Marburg and Ebola?

Having a better ear for and adjusting to the needs of the DoD is also one reason why I believe Tekmira will continue to have the upper hand over competitor Sarepta with its DMD distractions and an investor base that could not care less about the Marburg program.  Although the targets were explicitly not disclosed, I wonder whether there are host factors one could target and which would be beneficial for various viral applications.  This would be an alternative to trying to come up with target sequences that are conserved across the viruses.

Finally, serving as yet another example of illustrating the value of biodefense contracts for platform technology companies, Tekmira disclosed plans that a first clinical trial with a lyophilized SNALP formulation, TKM-EBOLA, is planned for the first quarter of 2014. 

Tekmira Sets New Standard in Ebola, Widening Reach of SNALP Delivery

When a small biotech company from Canada trading at less than 1CAD announces plans to list on the Nasdaq, one ought to pay attention, especially when this company is well known for its fiscal conservatism. The recent publication (Geisbert et al.: Postexposure protection of non-human primates...) on the treatment of Ebola infection by Tekmira and collaborators at Boston University and the US Army Medical Research Institute of Infectious Diseases (USAMRIID) in the prestigious journal The Lancet indeed supports that their optimism of good news to come out over the next couple of months is well-founded.

By demonstrating the complete protection of non-human primates, aka monkeys, from an otherwise lethal dose of Ebola virus using the company’s SNALP delivery technology, this paper sets a new standard in the development of anti-hemorrhagic fever virus treatments. More than the potential direct financial value of such a drug in the form of further biodefense funding for the benefit of SNALP technology development and long-term prospects of stockpiling contracts, the real significance of this publication is that it confirms SNALP, developed and owned by Tekmira, to be the industry-leading systemic RNAi delivery technology where successful non-human primate data have become routine, and furthermore extends the reach of this technology from targeting genes in the liver and solid cancers (already a plethora of applications here) to infectious and likely inflammatory diseases, too. Given that Big Pharma is well aware of the potentially revolutionary nature of RNAi Therapeutics and Tekmira’s current key enabling position in the space, it should only be a question of time that a bidding war for Tekmira will commence, if it has not already.

Ebola presents a biosecurity threat due to its deadly nature and the concern that it may be weaponized while no drugs can successfully treat it. Oligonucleotide-based approaches look most promising in the race to develop drugs against Ebola and other hemorrhagic fever viruses. AVI Biopharma in particular has staked a good part of that company’s future on winning biodefense stockpiling contracts from the US government, and as a result their antisense morpholino protein translation inhibitors have historically led the development. Such therapeutics aim to slow down the spread of the virus enough to give the immune system a chance to catch up and clear the virus.

In 2006, AVI and their collaborators around Dr. Bavari from USAMRIID reported results on a combo of morpholino antisense molecules against 3 Ebola genes (L polymerase, VP24 and VP35- targeting 3 different genes to address potential mutational drug resistance) in mice, guinea pigs, and non-human primates (rhesus macaques; Warfield et al.: Gene-specific countermeasures against Ebola virus...). In each case, 1,000 plaque forming units (pfu) of Ebola virus were administered, doses more than sufficient to essentially kill all the animals. In mice, somewhere between 50-500 microgram of drug (about 2-20mg/kg) administered either a little bit before or after the viral infection was enough to effectively protect the animals. In guinea pigs, 10mg of each oligo (~30mg/kg) also protected the animals well. Interestingly, even more protection was observed when the oligos were administered relatively late (up to 4 days) after viral infection compared to administration before viral challenge. While the pharmacologic explanation remains to be worked out, the ability to protect from Ebola relatively late after the infectious event is an important property of an Ebola antiviral.

While mice and guinea pig data are a routine component in the development of Ebola antivirals, success in these models by no means predicts success in primates, including humans, where the rhesus macaque model has become the gold standard. For this reason, 1,000 pfu of the Ebola virus Zaire were injected intramuscularly accompanied by a complex schedule of essentially daily drug injections (12.5mg to 200mg per dose) starting at day 2 before up to 9 days after viral challenge.

Initially, only a single antisense oligo against VP35 was tested which showed most efficacy in rodents. This proved to be ineffective with all 4 animals dying within 7 and 8 days post-infection. However, using the triple combo, 2 out of the 4 drug-treated animals survived the virus with one succumbing to a bacterial infection (= 50% or 75% survival depending on how you look at it). The most common side-effects were thrombocytopenia and an increase in liver enzymes, both of which are consistent with Ebola infection.

Until the Tekmira study, this has been the most advanced published study of an Ebola antiviral drug. One draw-back of the non-human primate studies, however, is that the treatment schedule does not allow for predicting how the triple morpholino combo will perform when given only after the viral challenge. Moreover, protection was not complete.

In unpublished follow-ups, AVI Biopharma and USAMRIID reported apparent advances in the potency of the morpholino chemistry (PMO vs PMO+). When rhesus macaques were treated with the new chemistry for up to 15 days, 75% of animals survived (I assume an n=4) ‘when the treatment period ended’. Clearly, more details are required to ascertain that this represented an improvement over the 2006 study, including survival data beyond the 15 day treatment period.

In light of these results, the SNALP-RNAi study sets a new standard for the efficacy of a prospective Ebola drug. This claim is supported by the fact that the model used in the Tekmira study was essentially identical to the one in the AVI study: rhesus macaques infected with 1,000 pfu of EBOV Zaire and treated with triple oligo combos against the L polymerase, VP24, and VP35 genes (note that any combination of siRNAs can be formulated in one step into SNALPs meaning that it is entirely amenable to rapid response applications). However, in this case the drug (2mg/kg siRNA) was given only after viral challenge, albeit starting already at 30 minutes after the challenge, either daily for 7 days or every other day for 6 days. Moreover, in the daily treatment regimen which is most similar to AVI’s study, all 4 animals receiving the SNALP-siRNA cleared the virus and survived while 2 out of the 3 animals in the alternate-day group survived.

Importantly, this was achieved in the absence of immune stimulation by the SNALP, which was very effectively abrogated by the 2’-o-methylation of the siRNAs. This is notable, because while the 2006 paper on SNALP-siRNA for Ebola in guinea pigs by the same group showed promising efficacy (>60% protection), it was clear that immunostimulation caused by the unmodified siRNAs and amplified by the SNALP was certainly a safety issue then (Geisbert et al.: Postexposure protection of guinea pigs...). Overall, the treatment regimen, 2mg/kg siRNA in SNALP every day for 7 days represents a very stringent test of the safety of SNALP and was very well tolerated, with minor increases in liver enzymes most likely attributable to the viral infection and not the SNALP-siRNA treatment.

It should also be noted that this study employed 1^st generation SNALPs which means that the 10 to 100-fold more potent 2^nd generation SNALPs as recently reported in Nature Biotech should widen the therapeutic window even more and is important when thinking about safety studies in human volunteers. It is clear, however, that while this study is an eye-opening proof-of-concept of SNALP-RNAi for the treatment of acute viral infections (at least those affecting SNALP-relevant cell types, phagocytic and hepatocytes in this case), any further development (objectives: pharmacology, biomarkers, testing SNALP RNAi at later time-points) towards biodefense stockpiling SNALP-RNAi for Ebola and/or an FDA-approved treatment for accidental needle sticks and sporadic Ebola outbreaks will depend on support by public funding agencies. It is difficult to imagine that this study can be ignored by the US government. The series of head-turning, top-quality SNALP papers in journals like Nature Biotech, Nature, Journal of Clinical Investigations, and The Lancet certainly won’t be ignored by Big Pharma.

Will “Target Protector” Technology Come In Play?

Last month, the Schier lab from Harvard extended their impressive work on the function of microRNAs in the zebrafish model of development (Choi et al.: “Target protectors reveal dampening and balancing of Nodal agonist and antagonist by miR-430”. Science 318: 271).

Antisense oligos to microRNAs are one method to experimentally infer microRNA function through loss-of-function analysis. However, this method becomes problematic when the intention is to study the biology of a particular microRNA-target interaction given the multitude of predicted targets for a given microRNA, typically estimated to be around a hundred per microRNA.

In order to overcome this problem, the investigators designed morpholino antisense oligos (morpholinos are quite popular with the fish community), which they termed “Target Protectors”. These are supposed to hybridise to a defined microRNA target site thereby blocking the function of the microRNA acting on that particular target site. This actually turned out to work quite well in the zebrafish and future publications should demonstrate the versatility of this new methodology.

The reason why I suddenly blog about it is because I had the pleasure today to listen to a presentation by Dr. Alex Schier in which it became apparent to me that there may be investor interest in commercializing the technology. This would complement the microRNA antagonist (“antagomirs”) approach currently pursued by the likes of Regulus, Rosetta Genomics, and Santaris, that aim at therapeutically blocking the microRNA. Given the many targets of a given microRNA, blocking them for therapy is therefore based on the belief that evolutionary pressure should have ensured that their targets are tied to a common biologic phenotype. Target Protectors, on the other hand, may be more specific in upregulating only one transcript that when de-repressed should be therapeutically beneficial. As such, one could imagine targeting the miR-122 target site of HCV, thereby avoiding potential side-effects due to inhibiting the most abundant microRNA in the liver.

It will be interesting to see who will take an interest in this emerging IP estate. It is also notable in this context that Rosetta Genomics, the microRNA target company, has previously mentioned their foresight to patent not only their discovered microRNAs, but also predicted microRNA target sites.

Update (23 November 2007): It has come to my attention that Target Protector has come into play indeed: http://www.stealthbiotech.com/index.asp