HBsAg Rules at International Liver Congress

Now that the HBV world has gathered extensive clinical experience with interferons and polymerase inhibitors (NUCs) and with the resolve to finally find a cure for a serious disease afflicting hundreds of millions worldwide, the hepatitis B surface antigen (HBsAg) has become recognized as the key determinant for treatment outcome.  This is being confirmed by trial after trial investigating combining the actions of both NUCs and interferons, either one after the other or together at once.

Some of these studies were presented at the International Liver Congress last week in Vienna (ILC2015).  Emerging from them are actionable HBsAg rules which can predict fairly well whether a patient will eventually seroconvert to (or at least lose) HBsAg.  No matter the excitement around CRISPR technologies, HBsAg seroconversion remains the gold standard outcome in HBV treatment in the foreseeable future.

These rules can be divided into pre- and post-IFN treatment onset.

In the pre-IFN setting, it is those patients that have below ~500 IU/ml serum HBsAg as a result of NUC treatment that will most likely respond to interferon treatment/immune stimulation with s-antigen seroconversion (see earlier blog entry).  Since NUCs alone hardly do anything to promote s-antigen seroconversion despite its dramatic lowering of viral HBV titers, it appears to be their slow impact on HBsAg levels ( 0.1 log per year HBsAg reduction) that has the synergistic effect with interferon: with HBsAg lowering you take off the foot on the immune brake, with interferon you step on the immmune gas pedal.

As such, HBsAg knockdown by RNA(i) Therapeutics would seem to do the same for interferons as NUCs do, only in a more rapid and potent manner.  Of course, both could be used concurrently as a run-in to IFN treatment.

However, once on IFNs (post-IFN onset), it is the relative HBsAg decline that has high positive predictive value in prognosticating who will seroconvert.  Of note, the HBsAg decline comes before any adaptive immunity can be detected.  This supports that HBsAg decline in itself contributes to seroconversion rather than it being a mere correlation.  In that setting, it is a 1 log decline in HBsAg the first few weeks after IFN treatment onset that separates the winners from the losers. 

It is uncertain to me, however, whether 1 log is a precondition to s-antigen seroconversion as the non-responders do not even come close to that (maybe 0.3log).  It is therefore possible that anything that pushes HBsAg below say -0.3-0.5log could have a dramatic effect on s-antigen seroconversion rates.

An RNAi Therapeutic for HBV used simultaneously with IFNs may therefore aim at helping IFNs to get to the  0.5-1log reduction threshold, and rapidly at that.

Ergo, there are now a number of obvious strategies that one can apply regarding the use of RNAi Therapeutics in HBV with various knockdown goals, both absolute and relative.  The exact strategy would depend on how the RNA agent is combined with polymerase inhibition/NUCs or immune stimulation. 

While a number of other HBV targets were reported at the conference such as core assembly and entry inhibitors, HBsAg (and HBV mRNA knockdown in general) lowering remains the most distinguished and the mechanism predicted to be most synergistic to existing treatment approaches.  As combination treatment is strongly predicted to be the future of HBV, HBsAg lowering should become a pillar of those treatment regimes.

Disclosure: long ARWR, looking for lower entry in TKMR.

Regulus Therapeutics RG-101 Continues to Have Potential in HCV Treatment Landscape

Having attended the International Liver Congress last week in Vienna, Austria, it has become clear to me that HCV is not going away soon.  Even in the US where progress towards its ‘eradication’ may be considered most advanced with about 1/5 of the known patient population treated last year alone (~250k), it will be an uphill battle to identify, treat, and pay for the millions more infected. Worse still, it is not enough to simply cure the existing pool of HCV patients, but also stop the cycle of re-infection (largely the result of injection drug use).

Treatment cost is one challenge and the current drug rationing approach leads to the counterproductive warehousing phenomenon where current medical intervention is focused on the patients with more advanced liver disease.  This is not only the hardest-to-treat population, but also allows the liver health of the previously less sick patients to deteriorate.  This obviously makes little sense also from a pharmaco-economical perspective when getting rid of HCV early on has now been shown over and over again to dramatically reduce HCV-related cirrhosis and liver cancer.

Another problem is the fractured treatment landscape that exists for the various patient populations (split up according to fibrosis/cirrhosis stage, genotype, co-morbidities, the rapidly growing concerns around drug-drug interactions, available/accessible meds etc) making it difficult for even the learned gastroenterologist to keep up with the latest developments and putting HCV treatment practically out-of-reach for the general practicioner.  

12-24 weeks remains the standard drug treatment duration with docs worrying about shorter treatment regimens being sub-optimal.  A triple regimen by Gilead after 4 weeks of treatment merely achieved a 27% SVR12 which in the DAA world is practically synonymous with a cure.  The best shot at shortening treatment duration may therefore come from Achillion with 6 week of DAAs achieving high cure rates in 'easy' patient populations.

While interferon is on the way out, it could make an at least transient comeback for the hard-to-treat genotype 3 where, in addition to cirrhosis, cure rates with the all-oral DAAs low (60-80%).

To sum it up, the liver community has expressed multiple times at ILC2015 that a short-acting pangenotypic regimen is an important goal in the development of new medications for HCV.   And if they paid attention at the oral late-breaker on Saturday, RG-101 is poised to play a critical role in filling this unmet treatment goal due to its long duration of antiviral activity following administration, regardless of genotype.

RG-101 update: more relapses, but thesis intact

The clinical investigators of Regulus Therapeutics presented a 20 week update on the phase I study in genotype 1, 3, and 4 patients with good to moderate liver health.  28 patients received study drug RG-101, 4 placebo.

At the primary endpoint on week 8 (reported in earlyFebruary), slightly more than half (15/28) of patients treated with RG-101 were below the level of quantitation (BLOQ).  According to the company, most of them were not only BLOQ, but undetectable (by sensitive PCR) at that.  This is a remarkable feat given that the GalNAc-conjugated phosphorothioate antisense molecule had been only given once.

According to the latest update, half of those patients eventually relapsed (7-8 depending on whether you count the patient that was lost to follow-up), most of them shortly after week 8.  Although the relapsers are slightly disappointing as in the short-acting DAA world undetectable virus for 8 (or better 12, SVR12) weeks following cessation of treatment is more or less equivalent to a full-blown cure.

Of course, the prognostic rules for a long-acting agent like RG-101 with a slower onset of antiviral knockdown ought to be different.  The notion that the viral rebounds were simply due to waning drug levels in the liver (and not due to viral escape mutations!), was supported by the biomarker analysis in the healthy volunteer part of the phase I study, also presented at ILC2015, where the trough in viral knockdown (~day 28) more or less coincided with maximal total cholesterol lowering as a predicted by miR-122 biology.

Of note, there was no apparent benefit of increasing the dose from 2 to 4mg/kg which was consistent with preclinical evidence that showed a declining liver/kidney drug ratio at 4mg/kg and maximal cholesterol lowering at 2mg/kg in humans.  This indicates that ASGPR receptor binding becomes saturated when too much GalNAc antisense is given at once.  The increased drug liver concentration at higher concentrations observed in earlier preclinical studies probably indicate uptake in non-productive compartments of the liver, including Kupffer and sinusoidal endothelial cells.

  

The hope is that with a second dose of RG-101 28 days after the first shot, maximal viral suppression can be maintained for at least another 4 weeks to stave off any viral comeback as seen in the single-dose study.  This is supported by both the healthy volunteer part of the study and the chimeric PXB mouse experiments presented which showed that such a second dose not only maintained drug potency, but in fact led to a step-up in efficacy.  I am therefore optimistic that with 2 doses of RG-101 monotherapy alone ~50% cure rates can be achieved in patient populations similar to that in the phase I study.

This, however, is not even the goal.  The ultimate goal would be to establish a simple pan-genotypic 4-week treatment regimen.  Accordingly, the combination of RG-101 with a DAA(s) (sandwich regimen) can be expected to result in a very, very deep viral knockdown by week 4.  At this point, a second shot of RG-101 would be administered to give the immune system another 6 weeks or so to finish off the virus.  I believe a very realistic scenario, and depending on which patient populations you are looking at a very compelling alternative to current HCV medications and those in development. 

PS: Open questions

Unfortunately, given that cholesterol lowering was still close to maximal at week 8, it would have been comforting to show a mutation analysis from the clinical study to confirm that viral relapse was not explained by the virus successfully developing resistance against RG-101.  This is such an obvious question that one wonders why Regulus did not present the data (yet). 

Another question mark around the current data set is why Regulus is not disclosing the differential effect of RG-101 on good and bad cholesterol and instead is reporting total cholesterol lowering.  The study investigator said that they are still analyzing the data from the multi-dose healthy volunteer study  (months after completing dosing???) and are planning to publish those.  I am mildly optimistic that this could unexpectedly bring to the fore the cardiovascular potential of miR-122 targeting, although in this case (à mostly chronic treatments) the liver cancer concern around miR-122 inhibition may be more valid than it is for the 2-shot HCV treatment goal. 

Absolute HBsAg Levels, Not Percent Decline May Be Goal of HBV RNAi

I just walked into an early morning session at the International Liver Congress in Vienna and may have learned the most important nugget of information regarding RNAi treatment for HBV.

During the discussion in the ‘Banishing B’ session, Dr. Joerg Petersen referred to clinical analyses to be presented this Saturday showing that patients who achieve HBsAg levels of less than 500, or even better 200 IU/ml and then are given an interferon have a much better chance of eventually losing HBsAg, i.e. being functionally cured, than those that do not.

Dr. Petersen has been involved in clinical studies combining nukes and an interferon, and data presented at last year’s American Liver meeting (AASLD) showed that 9% treated with the combination for 1 year lost HBsAg in the year following combination therapy compared to just 2.8% receiving interferon alone.

Note that currently only interferons are thought to give you a chance of functionally curing HBV by medication and that HBsAg seroconversion attributed to interferon treatment may occur in the years after cessation of interferon therapy.  It will therefore be important to see whether this increase in seroconversion with combination treatment continues to hold up in subsequent years. The nukes are ‘only’ thought to protect the liver from the ongoing damage from HBV replication and may have to be given indefinitely.

I am not sure whether the new insight comes from a fresh clinical study or are the result of a more detailed subgroup analysis of the previous study.  Regardless, it hammers home a message that can be heard again and again, even more so in Vienna this year than at the London meeting last year: it is all about HBsAg lowering.

Implications for HBV RNAi

The implications of the new analysis for RNAi approaches for the treatment of HBV is obvious: use RNAi to get HBsAg below the threshold. In other words, it may be less about getting a magical 1log knockdown, but more about getting patients below 500IU/ml. 

For this, Dr. Petersen recommended using 2 nukes so that the tiny and very slow HBsAg declines observed on nukes continue ( < 0.1 log per year). With RNAi agents, you would likely achieve this goal for a patient with an HBsAg at baseline of 1000IU/ml within a month, even with a single 2mg/kg dose of ARC520 from Arrowhead Research as reported last year. 

The absolute knockdown potency of an RNAi agent would determine which fraction of the HBV population fall within this sweet spot. With a 1log knockdown e.g. you could start RNAi with patients with as much as 5000IU/ml.

For the clinical development of RNAi agents for HBV, the first step will be to determine the HBsAg decline on a background of nukes (just what Arrowhead is doing right now).  For pivotal trials, an immunostimulatory agent such as an interferon should be added to the nukeà RNAi/nuke treatment regimen to finish off the virus.

Stay tuned to learn about the fold benefit of HBV cures in patients falling below the 500IU/ml threshold, and the 3-4mg/kg single-dose results for ARC520 from Arrowhead Research to be reported later this quarter (note to Arrowhead PR department: next Monday may be a good time to do so).