Alnylam’s Second-Generation GalNAc Data Impress

Earlier this week at the American Society of Hematology meeting in San Francisco, Alnylam presented first multi-dose knockdown data in humans for its second-generation GalNAc-siRNA conjugate technology (ESC-GalNAc).  Accordingly, an almost 60% knockdown was seen in the first (and so far only) patient dosed at 0.045 mg microgram per kilogram ESC-GalNAc siRNA with the knockdown yet having to reach its nadir.  Moreover, further knockdown benefits are likely to accrue with dosing beyond 3 times weekly as was the case in this study.

At these low doses, no remarkable adverse events were reported, including injection site reactions or increases in liver enzymes which were seen with first-generation GalNAc-siRNAs in humans at much higher doses.

Though the data from this phase I study of ALN-AT3 in hemophilia patients is still early and there will be gene target-dependent differences in knockdown potency, they do support Alnylam’s claim that second generation GalNAc is indeed considerably, as much as 50x more potent than first-generation GalNAcs.  Only last week I had speculated that it may be more appropriate to estimate 2^nd gen GalNAc to be ‘only' 5x more potent than 1^st gen GalNAcs.

I was wrong.

In non-human primates, ALN-AT3 had an ED90 of 0.5mg/kg with weekly dosing.  Based on the phase I data thus far, I expect the corresponding ED90 dose in humans to be a ~ weekly 0.075mcg/kg 0.075mg/kg for a ~7-fold improvement in potency in humans versus non-human primates, a relative potency improvement that is not seen for first-generation GalNAcs.  Add to this the inherent 5x improved intra-species potency of second- versus first-gen GalNAc (Nair et al 2014), it is indeed possible that ESC-GalNAc enjoy an up to 50x improved potency advantage.

In other words, it should be possible to now achieve highly potent knockdown with a once-monthly dosing regime and sub-1 ml subcutaneous injection volumes. 

The value of this for the ALN-AT3 program, which harnesses a unique mechanism in an otherwise crowded hemophilia field which in general is moving towards less frequent dosing/infusion regimens and less immunogenic molecules, remains to be seen.  The slow enrolment pace of the phase I trial, ~5 hemophilia patients enrolled in 6 months, gives cause for commercial concern.     

The competition

While the improved potency is great for Alnylam, the gene knockdown competition is unlikely to yield the liver to Alnylam.  Highly potent knockdowns are possible with other platforms by Arrowhead, ISIS, Tekmira, and possibly Dicerna, too. Beyond route of administration [subQ for Alnylam GalNAc, ISIS (GalNAc-) ASOs, Arrowhead’s single molecule DPCs, and Dicerna GalNAc versus intravenous for Arrowhead’s 2-molecule DPCs and Tekmira’s SNALP LNPs], differentiation will come from safety/tolerability and dosing frequencies.

The next datapoints in this regard will come from the R&D Day by Dicerna early next week (Dicer-substrate GalNAcs) and the 4mg/kg phase I data from the immensely exciting anti-miR122 program by Regulus Therapeutics (employing GalNAc-antisense chemistry) in January 2015.

Detailed Genetic Modeling Triggers Change in Hemophilia RNAi Target Gene Selection: ALN-APC Out, ALN-AT3 In

This week, Alnylam presented data for its hemophilia program which aims at providing particularly those patients that have developed antibodies ('inhibitors') against the recombinant factor VIII and IX standard-of-care with a treatment alternative.  After ALN-TTR for TTR amyloidosis, this is the second of the two programs the company wants to focus its internal development resources on.  The presentation, however, showed that the timelines have been delayed due to a change in target gene selection.  So with an IND planned for this program in 2013, it thus appears that the original 5x15^TM, which stated that the company wanted to move 5 clinical candidates into late-stage development by 2015, is more and more turning into a ‘1 out of 5’- if all goes well.  What a difference 18 months can make!

Sarcasm aside, the reason for the change in target gene selection is due to modeling the impact of various degrees of gene knockdown on the desired biological outcome: a 50% knockdown of antithrombin (AT) goes much further in terms of thrombin generation (the biomarker used in the study) than a 50% knockdown of the target gene in ALN-APC, protein C.

The type of detailed genetic analysis behind this realization is actually a very important one that companies should think more about when selecting RNAi target genes.  All too often, target gene selection is based on classical black-and-white gene knockout genetics which can be misleading.  Indeed, the VEGF component in ALN-VSP02 may be one of those. 

In the presentation, Alnylam further emphasized that ALN-AT3 utilizes a conjugate-siRNA approach amenable to subcutaneous administration.  Although the hemophilia community is very familiar with the intravenous route of drug administration, the company essentially claims that this is ‘a highly preferred mode of administration in the setting of hemophilia’.  

Really?  In any case, adopting GalNac-siRNA conjugation as an alternative to the gold-standard SNALP delivery would also make strategic sense for Alnylam.  Alnylam has become overly dependent on Tekmira’s technology to the extent that it apparently/allegedly felt compelled to mis-appropriate the technology which is subject to a high-profile ongoing litigation.  Not a good position to be in when the supplier (and owner) of that technology could pull the plug any day.  

On the other hand, the scientific evidence, particularly the shallow dose response in non-human primates which suggest that antisense-type large amounts of drug would be needed (3-10mg/kg), suggest that just maybe GalNac conjugation is not ready yet for prime time.   .