Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
BMRN -
BMRN -
BMRN
Disease Charities misfocus (one more follow up to my BMRN AdComm review post)
More reflection on self disclosure of AE
BMRN – Comments on AdComm relevant to Drisapersen (I’ll comment later on portions that appear relevant to Sarepta).
First, the no surprise items:
a) The rate of thrombocytopenia (the most significant adverse event category) – well known as the primary concern for Dris (if you were paying attention – although most SRPT bulls focused almost exclusively on proteinurea).
b) The severity of the thrombo – again well known since patients were hospitalized. But the exact details are, of course, new.
c) The late onset of thrombo – again not news. The severe thrombo events all occurred in the extension study.
d) The rate and severity of proteinurea – not generally news (but see one comment below).
e) The rate and severity of injection site reactions – not generally news since the AE levels were known. (That said, the images of the severe cases are fairly unpleasant even if they aren’t technically news).
f) The FDA commentary was pretty much as expected – e.g. the slam on anything without strong clinical efficacy data is completely in line with their prior guidance (although the Sarepta bulls bizarrely overlook it).
The new bad items (to me at least):
a) The sudden onset of the severe thrombo – which means it will be difficult to prevent. This is bad (as I’ve noted for some time – a key piece of the review will be how the SAE can be handled)
b) The case of glomerulonephritis
c) The duration of the skin reactions (literally years in some cases)
d) The fact that the long duration single arm patients that have done so well were healthier than normal. (Note that this very likely to be an issue for Sarepta as well – as I’ve noted before their lung data indicates they were substantially healthier at baseline than typical.)
e) The blood vessel inflammation in monkeys (although I haven’t yet gotten to any section discussing it in humans).
f) The FDA asked for CINRG natural history data that was not delivered with the NDA (and the FDA couldn’t get it either).
g) The FDA has been pushing for RCTs that are 2 to 3 years long. (My comment is that this is probably the right duration – 1 year is too short given the disease course.)
The new good items (again, new to me at least):
a) The Cumulative Distribution Curves for the 6MWD RCTs look fairly repeatable, which lends credibility.
b) The totality of the CK data is fairly strong – although, as the FDA notes, the CK can be fooled. But as some know it is the first thing I look for because it is a much more standardized (and thus reliable) assay than the various dystrophin markers. E.g. It is, in fact, often the first indicator that a patient has DMD.
And the ‘who knows’ items:
1) The one off events – i.e. the MI, the GI blockage and the thrombosis – are all known to be in areas more common in DMD. Thus it wouldn’t be possible to determine the drug dangers (if any) in this area without a much bigger RCT.
2) The dystrophin assay data.
Note that the reveals are generally of the magnitude that I expected - so I am a little surprised at the twitter reaction by the biotech cognoscenti. The only item where I am somewhat disappointed in Biomarin for not being more forthright is the sudden onset nature of the severe thrombo - the rest is just par for the course unless we start expecting every company to publish a 200 page self flagellation piece. (And note that even the Clinical Trials site doesn't allow for that level of self disclosure).
BMRN SRPT
FGEN
FGEN AKBA etc - Competitors in HIF anemia space:
FGEN -
FGEN - CHF data points and brief update earlier renal comments:
I've been trying to understand the CHF data - such as it is. The problem is that there isn't much of it - and what there is has problems of confounding or being otherwise weak. Nonetheless, just as a FWIW below are some of the more interesting papers on the topic
www.bloodjournal.org/content/bloodjournal/111/6/3236.full.pdf[tag]Somatic inactivation of the PHD2 prolyl hydroxylase causes polycythemia and
congestive heart failure[/tag]
Summary - animals without PHD2 develop CHF. But it is difficult to separate from the possibility that CHF is caused by thicker blood.
Hypoxia-Inducible Factor-Dependent Degeneration, Failure, and Malignant Transformation of the Heart in the Absence of the von Hippel-Lindau Protein
Summary - This shows that VHL (which is close to HIF in the networks) causes CHF independently of any blood issues. But the problem is that VHL is not PHD2 - even if close by. E.g. this VHL mutant produced large numbers of tumors, not something HIF or PHD does.
There are also some questions raised in at least one of the larger human PHD/HIF familial mutations studies that indicate somewhat higher incidence of CHF - but weak (at best) and again confounded by the thicker blood.
All told - as a working assumption I'd assume somewhat more CHF and/or more worsening with PHD inhibition or HIF raising. But probably not a huge effect since it does not show up as a huge issue in families that have over-expression of HIF. (Note: CHF NYSE Class 3 and 4 are excluded from the majority of the ongoing ph3s. Perhaps all given that ClinicalTrials.org often doesn't list all inclusion/exclusion criteria. But, like the above papers, there are multiple possible interpretations of that - e.g. trial designers typically want to avoid patients that are too sick in order to avoid confounding issues like hospitalization.)
Renal comment update: As I've noted before I think it is likely that in the non-dialysis patients there will be more hyperkalemia in the treated patients, but that that may not actually be, in aggregate, a bad thing since often hyperkalemia is a sign of kidney remodeling that eventually leads to slowing disease. (all renal treatment drugs commonly used today have this signature to one degree or another). The mildly interesting update is that AstraZeneca (a Roxa partner) just purchased a hyperkalemia drug company.
All FWIW - and more in the vein of having a framework to understand the data when the trials complete.
FGEN
TRVN - Notes from the Oct 2015 R&D Day by Trevena:
1) TRV027 in AHF:
a. New facts: not too much, but for me (I track only some of the cc’s): Ph2 study planned to complete in 2Q16; the aggregate endpoint for the ph2b is Dyspnia 5 day VAS AUC, 30 day mortality, 30 day rehospitalization for HF, 5 day worsening hearth failure, and length of hospital stay; and they provided the powering assumptions for each component.
b. Color: They explicitly mentioned in the charts that they understand the phenotype variability as an issue (what I have referred to as disease heterogeneity). My comment is that their recognition of that issue is a big risk reducer IMO.
2) TRV130 (oliceridine – IV opioid):
a. The hypoventilation was defined to be persistent lowered respiratory rate, lower respiratory effort or lowered blood oxygen – but what counted as lowered rate etc was physician choice. Note that CO2 was not a criteria for judging hypoventilation – but in answer to a question they said they were looking at possibilities for ph3.
b. They said they are looking at some potential additional trials – in patients for whom the advantages of their drug may be even better. It wasn’t precisely clear to me how this would fit into the ph3 program.
c. They haven’t yet talked to the FDA about the endpoints (presumably since they are still refining the analysis etc)
d. They have no plans to test their drug against Fentanyl (despite the fact that it is the drug of choice in the PACU – which is first place pain is controlled. So I do not really get this. E.g. how is the switch done today for Fentanyl to morphine as patient moves from PACU to normal hospital floor. Won’t they need data on the interaction between Fentynal and 130 as the switch is made?)
e. Trial site variability – they clearly worked hard to find trial sites which would allow specifying the surgical protocol (most hospitals specify the acceptable protocols – and thus there is large variability between hospitals).
f. Color: they had as one of their presenters the organization that designed their soft tissue trial and he clearly understood problems with trial design that I have never before seen any biotech directly address – e.g. increasing the “signal to noise” by reducing treatment variability and designing metrics that are more repeatable. Even more than in TRV027 they seem to understand what causes trials to fail and directly address them.
3) TRV734 (oral opioid)
a. From comments it is clear they just started shopping for a partner.
b. They do not intend to try to find their own anti-abuse system – since, for instance, many potential partners will have their own personal favorites.
P-Hacking - techniques to find it in papers:
The following is an interesting example and explanation of the technique:
http://blogs.discovermagazine.com/neuroskeptic/2015/11/10/reproducibility-crisis-the-plot-thickens/#.VkNzqoo77CQ
A thesis on synergies in IO – and particularly in relative sequencing and timing of immunotherapies. This is a further expansion of posts I have made before and covers not just the current technology/state, but also a discussion of the leaders in this area and other notes. As a thesis it will undoubtedly evolve but I wrote as a way to organize my thoughts and it provides a framework as I collect related data.
First, an argument from first principles - a comparison of IO vs traditional chemo oncology:
a) Chemotherapies are generally not strongly synergistic, and are not strongly dependent on precise timing/sequence. I would suggest that this is a fairly obvious outcome of the fact that chemo efficacy and synergy is largely driven by mutation – i.e. a random process. There might be an optimal timeline, and there might be synergies, but in general the amount is fairly modest because whatever synergy there is is both driven and masked by population and timing randomness (e.g. mutation events).
b) In contrast the immune system is a 'designed' system – and we know for certain that some effectors are strongly synergistic (e.g. vaccine adjuvants are designed for this) and some are well understood to be designed to happen in a certain sequence (e.g. innate then adaptive) and thus likely synergistic with the proper timing/sequence. This in no way is to say that there aren’t random factors at play too (e.g. epitope spreading is driven via intentional/designed mutation), but instead is to point out that with the right agents and/or the right timing the synergies are likely to be much stronger than in traditional chemotherapy combos.
Second, the data (such as it is):
The published IO synergy data is remarkably sparse in the sense that it is almost all in one corner of the trade space – i.e. jam together two adaptive phase (i.e. primarily acting on T-cells) checkpoint inhibitors. There is much less on checkpoint combos with innate actors (e.g. ADCC or KIR (see Innate data)), and even less data on optimization of sequencing/timing. Given the fairly large amount of data on checkpoint combos I won’t provide any links, but some of the more interesting data on other combos is:
Innate’s Product Pipeline section
Power of sequencing – Trastuzumab with CD-137 agonist
TME: ADCC and Checkpoint Sequence Optimization
Overview of potential ADCC synergies
Primary players:
Most of the data and on-going trials that are NOT of the simple jam-together-two-checkpoint is isolated to just a few players:
The most prominent researcher (by far) is Holbrook Kohrt from Stanford (his name is on well over ½ of all recent papers/posters that are on the topic of synergies outside of jam-two-checkpoints-together).
The companies that seem to touch some aspects of this (besides Innate) are a little uncertain but: BMY? (see the trials referenced in the sequencing vs TME link), TEVA? (they have a weak PD-1 which they smartly combined with Rituximab) and AFMD? (their ADCC actor combo with PD-1 appears to understand timing – probably because of Kohrt)
Commentary:
For the purposes of this post I would suggest that the oncology community has gotten so used to the *relative* unimportance of synergy and the *relative* unimportance of precise drug sequencing that they are more blasé about it than they should be. Further, to the extent that synergy is explored it tends to be either highly detailed mechanisms research (e.g. TME papers abound) or gross slam-it-together kludges that are not building on already extant knowledge (e.g. innate then adaptive). This is not to say that, for instance, TME research will not eventually help clinically – but it won’t be quick given that the TME is a time varying system of 100s of parameters that are explored by TME research 10 or 20 items at a time via snapshot.
Random notes:
Corollary – if correct timing/sequence is, in fact, very synergistic then combining two checkpoints in one drug likely to significantly suboptimal.
Random recent paper – on innate system guiding in the adaptive system: http://www.sciencemag.org/content/349/6252/aaa4352.short
AKBA
TRVN
PPHM
CETP inhibitors
FGEN AKBA - random notes and cites on HIF and diabetic nephropathy. I've noted for a while that it was difficult to sort out the effects of HIF promotion on chronic (vs acute conditions). After a fair amount of research I think that the common assumption that fibrosis (which is promoted by increased HIF) is largely synonymous with disease progression is not necessarily correct. At least for nephropathy. As a general rule it looks like additional HIF expression decreases the rate of progression:
Genetic difference that enables HIF correlated to nephropathy resistance
Overview (behind paywall)
But causes fibrosis:
http://www.nature.com/ki/journal/v84/n3/full/ki2013130a.html
BTW - I think it is a fair bet that HIF upreg induces hyperkalemia (it shows up in a lot of different trials although I haven't tried to track every non-dialysis trial) This is a fairly common occurrence with nephropathy treatments (even ACEs and ARBs) - so somewhat perversely this too is evidence. Of course this doesn't mean that the might not be negative consequences (there are trials that show reduced progression of kidney disease but still increase mortality).
SRPT - Another cut at the p value. Earlier I calculated the p value based upon loss of ambulation data (not a good p value), but with more time I tried to disentangle the individual patient data and run a p value with the entire set of patients.
However, before getting to that, some reminders:
a) small single arm trials with a small number of centers almost always look a lot better than historical. Often clinically meaningfully different - an illusion presumably caused by better SOC, better patient selection (and these were healthier patients than typical for their 6 MWD - e.g. see FEV/FVC data), ... . This is a bias that no statistics can remove.
b) Past behavior from Sarepta is that they appear to have used 'Difference of Means' to calculate p values - and it should be expected that that calculation would be very misleading for data like that in DMD. (Note that Sarepta is hardly alone in their apparent touting of 'Dif of mean' statistics when they know perfectly well the FDA will calculate p using a very different method).
Given #b and my p value calculation from the non-ambulatory patients I expected to find a p value not as good as they claimed. However their p value looks valid (at least as far as it is possible to tell given the difficulties of reading their individual patient data).
Why is the p value so much better than the non-amb p value and previous p values?
1) It is better than the non-amb p value because even comparing the still-amb patients the Etep patients didn't degrade anywhere near as much.
2) It is better than my past calcs because, of course, previously I was using Mercuri (the dataset that Sarepta themselves had previously chosen to compare to) as the comparator and now their data/claim is that P51 patients do *much* worse than overall Mercuri. Credible? Depends upon how well those patients match the Etep patients (baseline, SOC, ... this is a component of the post hoc issue mentioned previously in these threads)
With regards to approval... previously I would have said little chance. But if the P51 comparison set withstands scrutiny by the FDA (e.g. does it match Biomarin's natural history data and do the historical patients match other Etep baseline parameters), there are no SAE surprises and the FDA chooses to look the other way on problem #a above then it may be approvable.
That said, I am willing to double down on my earlier 'bet' at 1:1 odds that they are not stat sig on the ITT primary endpoint of any of their on-going phase 3s and further that one or more trends the wrong way. I.e. they cheaped out on their trials and the lessons learned that were evident today have not made it into their trials.
I am also willing to bet that in a head to head trial with Dris, Dris would have meaningfully more efficacy.
SRPT
SRPT
Shkreli