AASLD Abstract Continues to Support Promise of ARO-HBV, but with a Twist

Yesterday, Arrowhead revealed the abstract for data on its new HBV RNAi drug candidate to be presented at the big annual Liver Meeting in November (AASLD).  The data continue to support highly potent knockdown of not only the surface antigen (HBsAg), but also other HBV components.

Dose response issues

When Arrowhead announced initial first-in-patient knockdown data from its 3^rd generation HBV RNAi candidate ARO-HBV, I found it suspicious for them to only reveal data for the 100mg and 200mg cohorts although more data was available at the time.  Curiously, the new data seem to suggest that there was no apparent benefit from going higher than 100mg in dose, at least when dosed with the current monthly frequency. 

Whether there is significantly worsening safety with increasing dose remains to be seen.

HBeAg positive/negative dichotomy

Similarly (but not as earth-shattering for sure) to the history of a prior HBV RNAi candidate by the company before, ARC-520, however, Arrowhead was able to learn more about its drug candidate as they went along in the study.  In particular, the first dose escalation cohorts (100, 200, 300, and 400mg monthly x3) encompassed a mix of HBeAg positive and negative patients, with 13 out of 16 being HBeAg negative.    

When Arrowhead then decided the last 2 cohorts to be just HBeAg positive, a remarkable increase in drug response was observed: whereas it took the predominantly HBeAg-negative patients 70-90 days to achieve 1.5log reductions in HBsAg, it looks like the pure HBeAg-positive patients only take 30-40 days to achieve the same.

Ergo, the reason why no clean dose response had been seen in the first cohorts is best explained by the fact that they were mixed and at that small cohorts.

Trigger choice likely explanation

ARO-HBV comprises 2 RNAi triggers: one that targets all (intact) HBV mRNA (X trigger) and one that makes sure the surface antigen is hit (S trigger) even in those patients with integrated HBV.  These typically lose the corresponding DNA element targeted by the X trigger during integration.

In the early days of RNAi, there was some controversy about the usefulness of using 2 or more triggers against the same target in terms of knockdown potency and specificity.  When I first started to practice RNAi in the lab in 2003, it was my experience that when you combined a very potent trigger with a less potent one, the knockdown was less than with the highly potent one alone.  As a result, I am a firm believer in the concept of RNAi trigger competition.

In the case of ARO-HBV it means that in HBeAg negative patients that have lost the X-trigger DNA, there will be one sterile/inactive trigger somewhat blunting the potency of the active one.

HBV biology is certainly complex and downstream events could also account for final knockdown differences.  Accordingly, in the HBeAg-only cohorts  it appears from the early observations that the nuc-experienced patients that in previous studies corresponded to patients that had lost the X-trigger-targeted elements by study enrollment not only responded as well, but in fact slightly better than the nuc-naïve patients.  It will therefore be important to learn more about their HBV integration status at AASLD to confirm or disprove the competition theory.

Review of Recent RNAi-Related Developments

The last 2 weeks have been busy ones in the lands of RNAi Therapeutics.  Here, I would therefore like to offer my take on the most important developments.

Alnylam’s Primary Hyperoxaluria Drug Breezes Through Clinic

With 20 subjects on Lumasiran, a GalNAc-RNAi targeting glycolate oxidase, for a median of 7 months, the investigational treatment for type I primary hyperoxaluria (PH1) ismore and more looking like a solid drug.  If the data presented last week at a Paediatric meeting in Turkey hold up in another 30 patients, it should fly through approval onto the market by the end of 2019.

While the relatively small numbers in each cohort has caused a bit statistical noise, it seemed like 3mg/kg monthly gives you an advantage over 1mg/kg monthly or 3mg/kg monthly in terms of the key biomarker urinary oxalate.  Interestingly, the company has chosen 3 times 3mg/kg monthly as the loading dose regimen in the now ongoing pivotal ILLUMINATE-A trial to be followed by dosing every 3 months. , the initiation of which the company has just announced.    

The core double-blind portion of the trial runs for only 6 months, so I am a bit puzzled how much the trial can inform on the value of the loading dose regimen and whether quarterly maintenance doses are sufficiently effective.   In any case, the trial design apparently was blessed by the FDA (incl. the urinary oxalate lowering primary endpoint), so surely they will know (e.g. based on preclinical animal data).

Regarding the competition with the primary hyperoxaluria drug candidate from Dicerna, the approximately 2/3 lowering of urinary oxalate and essentially all patients getting into a range considered safe and an unremarkable safety profile sets a high hurdle for the upcoming data presentation of DCR-PHXC at ASN later this month.

Arbutus HBV-RNAi Drug About to Die   

Yesterday, Arbutus provided an update on its LNP-enabled RNAi drug candidate for the treatment of chronic HBV (ARB-1467) and it does not bode well.  Similar to most other drug candidates attempted and barely tried, ARB-1467 should be canned in the not-too-distant future. 

In a combination study testing ARB-1467 in chronic HBV patients on nucleoside reverse transcriptase inhibitor tenofovir, probably only 1 in 6 patients hit a pre-defined criteria in terms of HBsAg lowering that would allow the patient to then receive immune booster PEG-Interferon.  The not very forthcoming revelation and the fact that the data was mentioned even after dropping another shoe that its small molecule HBV RNA destabilizer is also biting the dust makes it clear that not even the company sees much value in ARB-1467.

What an utter disaster this company has been since its inception as a HBV solutions entity.

Arrowhead Early HBV Deal Further Validates

Last week, shareholders in Arrowhead Pharmaceuticals got a lesson in biotech investing as it shares have given up ~1/3 of its value following a nice deal with Johnson&Johnson for the leading HBV knockdown asset in the space, ARO-HBV.

The deal gives Arrowhead a solid $250M in financing, including a $175M upfront fee and a $75M investment in company stock at a premium of ~50% to current trading.  While this avoids an imminent dilutive secondary share issuance- the company had been running low of cash- the ‘up to 15%’ in royalties from future sales of early phase II-stage ARO-HBV is clearly disappointing.  The relatively low royalty despite of it having confirmed strong HBsAg knockdown in patients can be explained with the fact that Arrowhead has avoided bearing the full risk of having to show proof-of-concept of HBV control.  This despite of having earlier paraded results obtained with an earlier DPC-based RNAi candidate (ARC-520) as strong evidence that RNAi can achieve such immune control.

While Arrowhead got punished for playing it safely, it can now focus on commercial home-run indications with highly validated targets and transition to being a broad platform company with a wholly owned attractive lead candidate in ARO-AAT.  Indeed, Arrowhead may emerge as the leading RNAi company as Alnylam, with a market cap ~8x that of Arrowhead, is still burdened by late-stage RNAi drugs with suboptimal specificity.

While humiliating, this serves Alnylam as a good reminder to also value innovation by the direct competition (note: specificity-enhancing technology now adopted by Alnylam has been available for a decade and could have been had for peanuts, e.g. from now defunct RNAi play mdRNA/Nastech).  Fittingly, today long-time archrival Silence Therapeutics issued a press release indicating that Alnylam is playing hard-ball with regard to Silence-owned IP purportedly covering ONPATTRO.

My guess is that Alnylam would rather risk an injunction and deprive patients of a very good medicine than to submit to what appear to me legitimate demands of a competitor.

Akcea’s antisense drug for TTR amyloidosis gets FDA nod

Last Friday, Akcea’s TTR drug TEGSEDI (licensed from Ionis) finally got FDA marketing approval for patients suffering from TTR-related polyneuropathy.  As expected from the delay in the approval process, TEGSEDI’s approval is accompanied by hefty black box warnings highlighting thrombocytopenia and renal risks of the (non-GalNAc) phosphorothioate oligonucleotide.  

Furthermore, the REMS program calls for frequent blood monitoring, essentially negating the claimed at-your-home convenience advantage over competing RNAi drug ONPATTRO.  Add to the poor safety profile clear efficacy disadvantages and pricing at parity with ONPATTRO, ONPATTRO should clearly win out in the marketplace.

Having said that, the TTR community is relatively small and key opinion leaders with relationships to a given drug company tend to support the drugs from the same company.

Second-Tier RNAi Therapeutics Companies Starting to Deliver Clinically

Alnylam has had a 3 year head-start on the competition with GalNAc RNAi Therapeutics, but now additional companies are populating the field with actual clinical data.  While the knockdowns appear robust and the chemistries well tolerated, therapeutic utility and the potential of liver toxicity from off-targeting will have to be demonstrated in upcoming longer-term studies.

Dicerna clinically validates new target for primary hyperoxaluria

Dicerna used to be the leader in applying RNAi Therapeutics to primary hyperoxaluria (PH), an orphan genetic condition caused by dangerously high levels of oxalate in the blood.  Oxalate crystallization and deposition may then poison first the kidneys and then various organs leading to premature death.  Unfortunately, as its old LNP technology proved to be impractical, Alnylam's directly competing program targeting glycolate oxidase (GO) assumed the lead.

When Dicerna then came up with a GalNAc-based follow-up candidate (DCR-PHXC), it also chose a new target: hepatic lactate dehydrogenase (LDHA), a target which may have be applicable also for forms of PH beyond the most frequently diagnosed PH1 and for which GO is a suitable target.  Cynics may suspect that the new target was not chosen because it is the better one, but in an effort to provide differentiation over more advanced competition.  To wit, first-mover advantage can be significant in the orphan drug arena.

Perhaps due to its long-standing, deep relationship with the PH community, Dicerna has been able to recruit patients more quickly than I had suspected (in light of Alnylam's parallel recruitment efforts) in its first-in-patient study.  First data from this study (PHYOX) were revealed last week.  To their relief, LDHA knockdown resulted in oxalate lowering in all 9 subjects (8 PH1, 1 PH2) tested.

Following a single-dose of either 1.5mg/kg or 3.0mg/kg, at least 30% reductions in urinary oxalate were seen and all or almost all subjects got to urinary oxalate levels at or below 1.1mmol/24hrs/1.73m2, a level predictive of end-stage-renal-disease freedom.

More detailed data will be presented later this year where we will get a better sense in how the results stack up with those of Alnylam.  Earlier this summer, Alnylam reported 64% mean urinary oxalate reductions relative to baseline before following more than one dose.  Another area of concern is that it appears that 3 out of the 9 injections resulted in injection site reactions (ISR).  To see ISRs is not surprising per se, but the frequency appears to be on the high end based on early experience and could be a competitive disadvantage.  Similarly, no word was said about liver enzymes.

Given that Dicerna has been rushing the PHYOX trial without so much as testing multi-dose regimens, I expect the regulatory agencies to demand at least 2 dosing schemes in the upcoming pivotal trials.  A nightmare scenario would see Dicerna being sent back to phase II to test the safety of multi-dosing, a fate that Regulus Therapeutics had suffered before in Europe.

Disclosure: Over the last few months, I have been largely playing the volatility of DRNA- both on the long and short side.  Looking ahead to the end of the year, I would be hesitant of taking a substantial long position due to a market cap approaching $1B with the potential of negative surprises both on safety and efficacy when more detailed PHYOX data will be presented.  On the other hand, a deal (such as on the IND-ready mystery candidate) could provide a catalyst to the upside at least in the short-term.

Arrowhead sees best HBsAg knockdowns so far with new GalNAc candidate

When similar to Dicerna Arrowhead had to retool its lead programs with the GalNAc technology, one question was whether they could achieve knockdowns as potent as with DPC delivery which employed explicit cell penetrating chemistries.  It was therefore cause for celebration when the company reported -2log HBV surface antigen (HBsAg) knockdowns at the World Gastroenterologists Summit, better than with ARC520 and ARC521 before (slides here).  Also, the knockdown was observed regardless of e-antigen status, but that was to be expected with the new sequence design.

Intriguingly, following 3 monthly injections, the knockdown curves still kept coming down, making it almost look like HBV can be beaten into submission by the knockdown effect alone and without the help of an immune boosting effect.

Unfortunately, only HBsAg results from the 100mg and 200mg monthly dosing cohorts were disclosed even though the company had dosed 9 other dosing and patient cohorts.  Based on cohorts 8-11, it appears that best results may in fact have been obtained with the 300mg monthly regimen which will probably be disclosed in the run-up to this year’s AASLD meeting.

Similar to Alnylam in 2014, after showing impressive RNAi target knockdowns in the clinic, the wait is now on to show that the knockdowns will translate into therapeutic benefits.  Given the speed of the company’s execution in general, I would expect pertinent data to emerge from the HBV program in late 2019/early 2020.

Disclosure: similar to Dicerna, I have been recently playing the volatility of ARWR stock, getting long as it had been consolidating around the $14 level, then selling and going short into last week’s 50% spike to a $2B market cap.  With the puts I then sold ($21 and $20 September strikes), I have a ~10% cushion should the stock zoom past $21 by the September options expiration and my max gain is ~13%.  This is intended as a short-term trade

HBV RNAi 2.0

Gene knockdown, in particular RNAi and RNaseH antisense, holds great promise in the treatment of hepatitis B viral infection.  It is currently the only practical means to potently inhibit all viral gene products*.  It therefore is poised to become a cornerstone of future treatment regimens aiming at functionally curing HBV, an infection predisposing more than 200 million patients worldwide currently to liver failure and cancer.

Arrowhead- lessons learned

Unfortunately, the field took a big hit last year when HBV RNAi trailblazer Arrowhead Pharmaceuticals had to abandon its efforts due to preclinical toxicity resulting from its particular approach to releasing the RNAi triggers into the target cell cytoplasm (monkey deaths due to the DPC).

Nonetheless, after more than a dozen trials in WoMan, the company had learned a great deal about HBV and how to best tackle it by RNAi.  Chief among those lessons were the observations that RNAi can suppress viral genes, most notably the surface antigens HBsAg by sometimes more than 2-3 logs.  Moreover, in HBe-antigen negative and those HBe-antigen positive patients previously exposed to polymerase inhibitors (‘nukes’), Arrowhead painfully found that most HBsAg is derived from host genomically integrated HBV.  Consequently, RNAi trigger target sites placed downstream of the HBsAg ORF may be lost and RNAi rendered futile (ARC520àARC521 transition). 

Finally, consistent with the experience with nukes and interferons, it appears that RNAi treatment success (functional cure) should follow complex viral and host immune dynamics and while intriguing changes were observed in the clinic with ARC520/1 (e.g. new lower baselines following treatment cessation), it remains unclear how long an RNAi agent would have to be given.  This has implications for whether intravenous routes of administrations are practical or not.

The competition

Arrowhead Pharmaceuticals, however, has not been the only RNAi game in town developing HBV therapeutics.  Its main competitor in terms of scientific prowess has been Arbutus Biopharma (renamed from Tekmira after biotech wonder boy Vivek Ramaswamy of Axovant fame spectacularly raided the company in 2015).  Its lead RNAi candidate ARB-1467 comprises of 3 RNAi triggers which are formulated in lipids (LNP) and is given intravenously alongside steroids.

While I like the 3-trigger strategy for pangenotypic coverage and for minimizing the risk of the virus developing drug resistance (including by genomic integration), the more cumbersome intravenous route of administration- now reduced to short 2-week intervals in an effort to increase potency- and the steroids makes ARB-1467 uncompetitive in a world of more potent and less frequent simple subcutaneous competition.  The use of immune suppressive steroids, of course, in HBV patients is a dicey proposition and would also seem to run counter to the ultimate aim of achieving immune control of the virus.

Unless it turns out to promote a functional cure along with other agents in short order, say less than 6-12 months, ARB-1467 will likely end up as a science project without much clinical impact.

RNAi powerhouse Alnylam Pharmaceuticals meanwhile is the third RNAi company that has begun clinical development of an HBV RNAi agent.  Importantly, it has been the first company using a simple subcutaneous GalNAc-conjugate format, therefore positioning it to be useful even when more prolonged treatment will turn out to be necessary.

After review of the program, however, it appears that the company prematurely rushed the single trigger ALN-HBV into the clinic without thinking too much about resistance issues.  To start with, ~2% of tested HBV genotypes have mismatches with the trigger that mitigate targeting efficiency.  As a single trigger candidate, ALN-HBV will also have to be given alongside highly potent replication inhibitors (nukes) as one can easily see how ALN-HBV resistant genotypes would otherwise eventually take over.

ALN-HBV moreover targets a site downstream of the HBsAg ORF, around the DR2 repeat element with marks the integration hotspot that has bedeviled ARC-520 before.  While Alnylam has been going around claiming ALN-HBV doesn’t suffer from ARC-520-type issues, I would challenge them with two points:

1)     under selection pressure by ALN-HBV to maintain HBsAg expression and thus evade host immune detection, the virus may ‘choose’ to break up upstream the ALN-HBV target site without affecting the HBsAg ORF; 

2)      more troublesome, closer inspection of the very paper Alnylam points to for making its claim (Jiang et al. Genome Research 22: 593) and which analyzes HBV integration hotspots, shows that ~40% of DNA break points appear to be upstream of the ALN-HBV target site (compare ‘position 1600’ below).      

One can therefore easily see why the project leader behind ALN-HBV, Laura Sepp-Lorenzino, has recently left the company to join Vertex Pharmaceuticals.  One has to get the impression that ALN-HBV, just like ALN-GO1 have only been rushed to the fore as a front in order to keep a lid on their competition by creating doubt about the ability of Arrowhead Pharmaceuticals and Dicerna, respectively, to compete with juggernaut Alnylam. 

Thrilled to join @VertexPharma as Head of Nucleic Acid Therapies!!!

— Laura Sepp-Lorenzino (@LauraSepplore) September 23, 2017

 

It is yet another lesson that in drug development, a detailed understanding of the disease is as important as the technology used to tackle it.  Half-hearted side projects typically lead nowhere.

Lastly, I would be remiss if I did not mention the RNaseH antisense efforts by Ionis along with partner GSK, and those of Roche.  Ionis/GSK are not only developing an unconjugated fully phosphorothioated antisense molecule, which I believe has little chance of competing in the market due to predictable safety and potency issues, but also a more interesting GalNAc-conjugate version (IONIS-HBV-LRx).  Although I currently see RNAi ahead of antisense in gene knockdown in hepatocytes (potency, frequency of administration, and safety), the GalNAc-conjugate version potentially has the advantage of also being able to access the pregenomic RNA directly, while direct pgRNA cleavage by RNAi of this non-mRNA remains to be shown.  My prediction is that  while RNAi can to some degree access pgRNA, this is not as effective compared to its cleaving mRNA.  What all of this means biologically remains to be seen.

Arrowhead HBV RNAi 2.0

Therefore, after all the drama and competitive noise, Arrowhead is poised to recapture the HBV RNAi lead with its new GalNAc-based candidate.  ARO-HBV is poised to enter the clinic in the first half of 2018.   It is subcutaneously administered and involves 2 RNAi triggers that are claimed to cover the viral resistance bases, including HBsAg derived from genomically integrated HBV.

The company expects the agent to be used once a month or less frequently, an attribute valuable should functional cures take longer to emerge.

Given knowledge leadership in HBV gene knockdown and prior practical experience, Arrowhead should also be able to navigate through the HBV clinical development maze faster than its competition.  As can be seen from its resurging stock price, this view is also shared by an increasing number of investors. 

Arrowhead has paid for taking some short-cuts when it was compelled to push ARC-520 into the clinic to give it a shot of becoming a serious player in RNAi before it ran out of capital.  It took some risks and failed, but that failure could well be the soil from which future success will emerge. 

Disclosure: long Arrowhead Pharmaceuticals

 

 

 

 

 

 

* it is unclear whether RNAi can directly cleave pregenomic RNA, whereas RNaseH antisense should be able to do that.