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.
AVXS
IONS AVXS
IONS R&D Day -
Key points covered (and I cover some in more depth below):
a) Thrombocytopenia data updates - bottom line is that it is probably a class effect, albeit a nowhere near as severe as many (including myself) thought from the initial cc (see post to which this post is a response). Also significant interaction with thrombo that happens in natural history patients.
b) Lots of natural history data in SMA and TTR FAP/FAC, FCS. Also some regulatory discussion - and a key point of which is if the Cardiomyopathy subgroup of the FAP trial is good they hope for something from the regulatory agencies (no specifics given).
c) They do a comparison of the ALNY TTR drugs vs TTR-Rx (reminds me of elementary school playground - "Fight! Fight!"). Undoubtedly the ALNY drug used for FAP is significantly more toxic than TTR-Rx, but I don't fully accept their arguments for the comparisons to the ALNY FAC drug - e.g. the IONS chart, which was based upon the ALNY cc presentation earlier this month, shows 44% 'still on trial' vs 95% for TTR-Rx, but this doesn't really match the data on the ALNY charts (which, I'll acknowledge, are not entirely clear to me).
d) GCGR-Rx further dose ranging data discussed with effects on ALT and HbA1c. (Essentially they can get the same HbA1c but with less ALT rise by lowering the dose substantially. See #msg-103431870 for a thread discussing the 2014 results that forced this further dose ranging).
e) miscellaneous stuff - showed ANGPTL3 cholesterol subtype data, which showed that, unlike oral ANGPTL3 data, their drug does not worsen HDL. Also made passing mention that the ph2/3's for APO(a)-Rx that are planned to start later this year and next are not planned to use "outcomes" data as the primary endpoint and instead there would be other clinical data. And, further, there was reference to the regulatory agencies understanding the importance of this target.
f) Discussed new technologies: LICA for muscles (which would improve efficacy in muscle - my note would be improves targeting). And method by which they could actually upregulate expression of many targets with anti-sense because more translation appears to be 'designed' to be inefficient, so anti-ing the inefficiency can increase protein expression significantly.
Thrombocytopenia - key points:
Key caveat: It wasn't always possible to be sure whether they were talking about thrombo in FCS or thrombo in FAP. I've made my best estimate - but ...
a) No one died. And it appears no one even had any obvious abnormal bleeding, so all cases were caught via lab testing. (Note from previous cc - there were less than a 'handful' of severe thrombo patients across the two drugs)
b) Natural history, recently pub'd, shows that FAP patients naturally have some thrombo (e.g. in NH study 22% had Grade 1 at baseline, 8% had drop under 100k at some point over 50 months). And FCS patients also natural thrombo - "cyclical" in some patients. In contrast the FPL patients (a separate Ph3 trial using the same APOC3-RX drug as FCS) appear to show no thrombo in natural history - and none in ph3.
c) No thrombo in safety ph3 with APOC3-Rx drug in high trig patients (COMPASS trial) - even though about 2x as many treated patients as FCS trial (albeit less duration - but reminder that all the thrombo appeared at 3 or 4 months of treatment).
d) Although Stan repeatedly said "not a class effect" the data he gave all said it is - albeit a class effect that is much more mild and workable than, for instance, the effects seen with Drisapersen. The data that says there is a class effect is:
d1) they showed a table of thrombo seen in a large set of IONS drug trials (later appeared to be described as all completed ph1/2/3 trials) and it showed an outsized thrombo at the highest dose (a percentage incidence 4 or 5x the incidence seen anywhere else on the table).
d2) Stan mentioned that at higher doses of IONS drugs there is a mild "trend" to lower platelets (albeit no clinical).
d3) Current belief of IONS is that the thrombo is an effect of peak blood concentration (plus, based upon their presentations on natural history, disease that naturally has significant platelet variation).
e) One MD on TTR trial mentioned that none of his patients with thrombo on drug needed steroids. Just drug holiday.
APOC3-Rx (Volanesorsen):
A) Aside from natural history and thrombo data the most interesting note was an aside that Stan made that 'if the FAC subgroup of the FAP trial shows benefit that they will talk to regulatory agencies about ...'. Also noted that they are not expecting more thrombo events since everyone has now been on trial for a while.
B) In their recent polling of a PCS patient group they have been gathering the average number of pancreatitis attacks was 32. And 70% had had their gallbladder removed (an indication of the rate of misdiagnosis).
Random Comments - at this point I am not sure how many events have happened in the SMA ph1/2 since in this cc one of the MDs said no new events since January (presumably of this year), but at the shareholder meeting in June they said there had been one death and one permanent ventilation.
Different medical communities - and the very different cultures for levels of evidence. I have previously noted that the DMD community appears to be fairly disfunctional. (see the post to which I am responding). But it shows up in all medical specialty communities at some point (e.g. my post several years ago about cardiologists in denial about the diabetogenic aspect of statins).
But the recent AC vote on empagliflozin is probably one of the more stark.
http://cardiobrief.org/2016/07/05/prominent-cardiologists-decry-tepid-support-for-empagliflozin-by-endocrinologists/
FWIW my personal opinion on endocrinology in general, after reading hundreds of papers/abstracts, is that it is one of the more backward fields. Part of this is the difficulty (fast varying interconnected control loops is hard), but part is what is on show in the above link.
THLD -
THLD - Comments on hypoxia, IO and THLD:
1) Having now listened to the last THLD, a brief summary of the TH-302 history:
a) Did great at pancreatic cancer PFS - both in ph 2 and ph 3. Just missed OS (p just barely above 0.05, although since it was a large trial the HR wasn't great). But the FDA rejected any attempt to file an NDA.
b) The Japanese subset of the pancreatic ph3 does, in fact, look reasonable - although I'd want to see the Japanese data from the ph2's and Japanese ancestry data from RoW to get good confirmation.
c) The other ph3, in sarcoma, completely failed (and, unlike the pancreatic which had an RCT for Ph2, the sarcoma ph2 was a comparison to historicals - it looked great compared to historicals! <g>). Note that THLD was running this trial, Merck was running the pancreatic - see my comments on THLD management.
2) THLD management was all about the post-hocs in the cc. I tend to avoid listening to conference calls of companies with this kind of ignorance, so it has been a while since I heard anything so annoying (e.g. they blamed the failure in osteosarcoma on the control group doing so well). This, of course, makes everything else they say suspect IMO.
3) As for the combo's with IO, the story that is told in the TH-302 literature (that I've found so far) is a little different than the story told by Curran in the XNCR presentation. In the XNCR presentation he talks of TH-302 providing "hypoxia ablation", but the TH-302 data is just about the drug being a hypoxia triggered pro-drug. See this Curran paper, for instance. I would think there is an important difference between "hypoxia ablation" and "drug killing the hypoxic core". Perhaps an evolution in understanding. Worth tracking IMO given the apparently strong data in models - but still very much a work in progress.
http://cancerres.aacrjournals.org/content/74/19_Supplement/629.short
XNCR - Summary of the recent analyst day presentation:
a) Further clarity on items currently in trials:
1) Summary of 5871 (anti-B-cell) protocols of on-going Ph2s in SLE and in IGG4 disease. IGG4 is single arm, single site trial copying protocol of Rituximab trial and based on Responder Index (very similar to vasculitis index). Could have data at end of 2016, but guidance is 2017. SLE trial is 90 pts in RCT. Protocol is to give steroids, to allow taper off immunosuppresant, and randomize those who respond to this lead in. Trial is to provide either treatment or placebo until they lose benefit seen with initial steroids. (Separately note that they are also planning a SC trial later this year to check PK vs IV).
2) 7195 (anti IgE). Nothing new (historical note that they have not yet been able to shake their adverse events of frequent transient thrombocytopenia (in the recently completed trials the low loading dose did not prevent the later, larger dose from creating thrombo), but they have a trial planned to start Q3 trying SC) Note: in response to Q it was said that they have not yet found the mechanism of the thrombo.
b) Oncology pipeline updates - won't try to recapitulate all the details. But some of the things I found most interesting:
1) Unlike their event last year they did not mention anything about targeting regulatory T-Cells.
2) Their focus was on bispecifics that target two different checkpoints.
3) They mentioned the future possibility of a bispecific that antagonizes a checkpoint and agonizes an immune costim.
4) My comment: most of the preclinical IL-2 data didn't look spectacular. Yes, a bispecific that targets PD-1 and xxx produces somewhat more IL-2, but not clear it is above the noise threshold of the test (i.e. for every test they used a straight PD-1 inhibitor as a comparator - and the IL-2 effects of that comparator bounce around a fair amount)
c) Raw science - I found the first presentation on this by far the most interesting. Michael Curran spoke on the topic "Hypoxia Drives Tumor Immune Suppression and Immunotherapy Resistance" (starts about 1 hour mark) and there was, for me, substantive new information:
1) Background/assertion: some of the cancer types that are known to be unresponsive to IO (e.g. prostate and pancreatic) are believed to be unresponsive because they are hypoxic and the hypoxia inhibits T-cell immune response.
2) He presented material on a mouse model that spontaneously develops prostate cancer and the effects of treating the mouse with TH-302 (a drug that reverses tumor hypoxia) with various combos of IO. Pretty dramatic effects.
3) In passing he mentioned things that were interesting: A) showed data that PD-1 plus PD-L1 is synergistic, B) summarized the added tox of PD-1 agent plus CTLA-4 isn't worse tox, but is instead more frequent tox, C) mentioned that oxaliplatin is significantly more immune priming than the other platins.
4) Also, in passing, gave details about TH-302: A) failed all its ph3s but showed efficacy in Japanese cohort (well known that Japanese patients respond differently in NSCLC, but now I am curious about other cancer types), B) they are planning a combo trial of Ipi and TH-302 to start later this year.
Cancer IO and synergies with older, but still IO, treatments like vaccines and ADCC antibodies (e.g. herceptin):
I spent some time during ASCO looking for abstracts that paired ADCC drugs with some of the major IO drugs. E.g. PD-1 agents with herceptin or cetuximab. Amazingly there were none. It is odd to me that this is so ignored - the first human tests will be from small biotech with new ADCC drugs. E.g. AFM13 (an ADCC drug via bispecific) paired with a PD-1 or Margetuximab paired with a PD-1. This is despite a significant, pre-existing, set of approved ADCC and IO drugs.
FGEN
FGEN - some additional notes:
TRIL - and on the flip side, another recent paper:
IONS - thrombocytopenia further thoughts:
a) There is some chance this is a broad class effect and would spread to other companies as their ph3s get fully enrolled and time on trial increases (e.g. similar sudden onset thrombo in Dris (a different technology) and general, but less extreme, thrombo in others).
b) If, in fact, a patient died I don't see how Ionis avoids a broad stop-dosing hold for all systemic administration trials. If this lasts any length of time it damages the ph3s even if(!!) a solution is found.
c) The neurological trials (large pipeline of agreements with BIIB) *may* be safe. But I'd want to see the general (ITT average) platelet levels plus SAE thrombo data.
d) The R&D Day in July is likely to get cancelled - too much of their pipeline might have just gone 'poof'.
e) I'd be surprised if at least one of their partners for systemic usage doesn't drop them in the next year.
Note: I would say that I think that if anyone figures out what is causing this it is likely to be Ionis. They have data which others don't and are good at the science (just very very ostrich-like when it comes to adverse events).
IONS
TRVN
TRVN - TRV027 fails AHF Phase 2:
Trevena Reports TRV027 Did Not Achieve Primary or Secondary Endpoints in BLAST-AHF Phase 2b Trial in Acute Heart Failure
Business Wire Trevena, Inc.
53 minutes ago
????
KING OF PRUSSIA, Pa.--(BUSINESS WIRE)--
Trevena, Inc. (TRVN), a clinical stage biopharmaceutical company focused on the discovery and development of biased ligands targeting G protein coupled receptors, today announced that the company’s TRV027 failed to meet either the primary or secondary endpoints in the Phase 2b BLAST-AHF study in acute heart failure (AHF). The company expects to focus its efforts on its lead Phase 3 oliceridine pain program and its earlier stage programs.
“We are very disappointed that TRV027 failed to show the hoped for benefits to patients in the BLAST-AHF study. We will continue to analyze the data to further understand the outcome, but believe the study was well conducted and has answered the questions it was designed to test,” said David Soergel, M.D., chief medical officer. “We are extremely grateful to the patients and investigators who participated in the trial.”
“It is always disappointing when an investigational product fails to support a promising hypothesis in a clinical trial,” said Maxine Gowen, Ph.D., chief executive officer. “We remain confident in our Phase 3 oliceridine program, which has been granted Breakthrough Therapy designation by the FDA, as well as our earlier stage opportunities and platform, and we remain committed to bringing innovative new medicines to patients to treat serious medical needs.”
Data from the BLAST-AHF trial will be presented in a late-breaking trials session scheduled for 2:15-3:45pm CEST on Saturday, May 21 at Heart Failure 2016, the annual congress of the Heart Failure Association of the European Society of Cardiology.
About the Phase 2b BLAST-AHF trial
BLAST-AHF was a randomized, double-blind, standard of care controlled trial in 618 patients with acute heart failure. The study compared TRV027 (1.0 mg/hr, 5.0 mg/hr and 25 mg/hr) plus standard heart failure therapy versus placebo plus standard therapy. The primary objective of this trial was to evaluate the effects of TRV027 on a composite of clinically important outcomes: mortality, worsening heart failure, hospital readmission rate, dyspnea, and length of hospital stay. In this study, TRV027 or placebo were initiated after presentation to the hospital and then continued to be administered for a minimum of 48 hours and a maximum of 96 hours. Pre-specified analyses to identify populations that may respond best to TRV027 included segmentation by ejection fraction, systolic blood pressure, plasma renin activity, and glomerular filtration rate. The study methodology was published in the Journal of the American College of Cardiology – Heart Failure in March 2015.
FGEN -
FGEN
CNCE:
CNCE -
SRPT -
SRPT - Review of letter posted that drug should be approved:
http://www.biocentury.com/biotech-pharma-news/coverstory/2016-05-02/why-fda-should-grant-accelerated-approval-to-eteplirsen-for-dmd-a02
I tweeted a little on this letter this morning, but really the letter has enough concerning content about Sarepta that 140 characters, even a string of them, is not adequate.
So let me take some of the significant ones on in a form somewhat easier to follow than twitter:
SRPT 6MWD vs IONS Infant SMA -
SRPT – Comments on FDA Review material – with emphasis on their critique of historical registry data.
Before going into specifics I think it is first important to understand a key fact of an FDA Review – that an FDA negative review is actually a LOT harder when the NDA has essentially no information. Why? Because it is much much harder to prove a negative with the specific data if there are few specific data. (that is, for instance, why the legal system presumes innocence, not guilt). For example, imagine a company submits an NDA with just in-vitro data and a story about that MoA and how well that in-vitro obviously maps to in-clinic cancer results. All the regular posters on this board know that this mapping is highly unlikely to be valid, real world, logic – but they know it from past rates of failure. Y’all would, however, have a very hard time “proving it won’t work” since the actual specific proof requires knowledge no human has about off-target toxicity, alternate pathways, specific breakdown products, … . For those on this board who have spent time trying to educate newbies you know how hard it is for the bulls to understand this without the specific disproofs. (Interesting note that lately there has been a rash of tech start up millionaires making the mistake of assuming that biotech is easy with just a little modelling – but completely missing the point that modelling can never be any better than your knowledge of what is being modelled and we ultimately know very little compared to the areas they are used to.)
Ok – given that background, this FDA Review is better than the last one in many respects. But it was particularly better in, to the extent possible, citing specific issues in the comparison to historicals. Yes, the old FDA Review did make some general references to the fact that trial data almost always looks much better than historical data. But not much in the way of specifics – and what specific critiques there were in this vein were not well thought out (leaving room for Sarepta critique – some of which was accurate in piece part although it totally missed the bigger picture).
I applaud the FDA because it was undoubtedly very time consuming and politically challenging to get the raw data from the historical database orgs – and it provides a very valuable insight into the problems with historical registries and comparison to them. A road map for what orphan disease registries should avoid it they want to actually help understand and cure their particular disease. Below I’ve listed arguably the 3 biggest damning items with regard to using the data vs the Etep data, but in reality many more of the items are of value to other registry orgs:
a) General and very meaningful sloppiness about gathering, keeping and reporting data – some of the bigger examples given below:
a1. Losing patients without reporting that fact accurately (seriously): the historical data originally claimed only a small fraction of patients were lost to followup, but in reality about 40% were lost and many were replaced by entirely different boys. I.e. such sloppiness pretty much invalidating the natural history nature of the data.
a2. Healthier patients drop out – because they can take part in trials. See above issue.
a3. And, if those are too generic, the data itself was shown to be unreliable - A key part of Sarepta’s story was that only 2 of their kids lost ambulation by 4 years, but all of the historical patients did. FDA fairly conclusively proved that this wasn’t true since at least 2 of the 13 historical patients (and possibly/probably more) had 10 meter walk/run data so they clearly could walk. Furthermore some of those “LOA” patients actually had fairly good 10 meter walk/run data – implying a fairly high 6MWD. (the implication is that they just weren’t pushed to take the longer, more difficult, 6MWD). And, further, the FDA proved that for other kids the dataset was just plain contaminated. (see page 31 of addendum)
d.
b) Historical database baselines (and incentives) are different than for trials: In order for the comparison to historicals to be valid the kids have to be the same at baseline and almost certainly they were meaningfully different. I’ve noted before that Sarepta’s own data about FVC etc indicates the kids were significantly healthier at baseline that normal for kids of their 6MWD. The FDA’s data addition was a comparison of northstar data that indicated the kids in the Sarepta trial were meaningfully more functional at baseline but are otherwise following the same trajectory (see page 51 of addendum). (Note also that this brings up another point that I will get to at the bottom of this post.)
Last comment on the review: A subtle implication within the FDA Review is that they believe that the CINRG group has been more careful in creating and curating their DMD historical data and the FDA documents make it clear that they expect to get further CINRG data prior to the AdComm. I’d expect these to be further damning of the very tainted historical comparisons that Sarepta has been using. (Note that in January I expressed puzzlement about the inability of either Sarepta or the FDA to get the CINRG data – but given current FDA Review I think the explanation might be the political nature of the current fight and a CINRG desire to be much more careful with their data than is, apparently, typical in DMD in general and Sarepta in particular.)
Finally – a general comment: The thing that surprises me most about some of the biotech cognoscenti reaction to the FDA review is the belief that the 6MWD is not influenced by externals like motivation or test methodology or … . This gap, between my understanding of the universe and theirs, is so large as to be almost incomprehensible to me. I am hoping that theirs is just a misbelief that the FDA is saying ‘the disease course is influenced by encouragement’ – and they disagree with this. If this is the basis of their disagreement with the FDA I would still side with the FDA, but I can at least understand the disagreement. But I cannot comprehend a belief that a performance test is not influenced by encouragement/discouragement. And note that most disturbing of all about 6 MWD and encouragement/discouragement is that the initial test, used to decide admittance to the single arm trial, was based upon having a walk distance under a certain limit – so there was a strong motivation not to try too hard on the initial test, and this would make all future comparisons against that baseline look better than they really are (and probably explains a large part of the discrepancy with the Northstar data). (Obviously with an RCT this issue largely disappears – but in a single arm trial it is a HUGE protocol issue.)
IONS ALNY - in the recent ALNY cc on their TTR PN OLE they speculated on the reason for the hold on the TTR-Rx cardiomyopathy trial:
TTPH -