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.
>>Because the trial has multiple (non-hierarchical) primary endpoints, causing the alpha spend to be divided among them.
I always wonder why a big company would do this rather than do enough work ahead of the trial to try to figure out what the best single endpoint is.
>>I never put a lot oF faith in the strength of those
Justifiably so - the use patents usually don't hold up because there are often workarounds even if the patents survive challenge.
A Cmin patent might hold up better - it was certainly non-obvious as can be seen by a decade of struggles with dosing the drug.
Peter
>FGEN IP
Worth noting that Neff mentioned a new patent filing related to Cmin. Whether that might hold up or not I have no idea.
This isn't a great candidate for a biosimilar - the inherently high COGS will cut into margins, not to mention the start-up costs for a huge plant.
FGEN has also been burnt before on both kidney and liver fibrosis - the endpoints are hard, although you could now piggyback off the other NASH players. Those early FGEN trials were almost certainly significantly under-dosed as well.
For example, this was run at 5 and 10mg/kg:
https://clinicaltrials.gov/ct2/show/NCT00754143?term=fg-3019&draw=1&rank=6
>>I bet that 58% number varies a lot more drastically if separated by Large, Mid, and Small cap bios.
There was someone's (Feuerstein maybe?) Rule along the lines of a how a biotech with a market cap less than some amount ($100m?) had never had a successful Phase III.
Of course that partly reflects the market's accurate prediction of the (un)likelihood of success, so causality runs both ways here.
Here is the perfenidone NEJM disclosure of the statistical model FGEN also used:
The slope analysis is a mixed model repeated measurement analysis with linear time effects for patients. Specifically, the mixed model of change in FVC volume uses time-by-treatment, age-by-sex, height-by sex terms as fixed effects, and patient, time-by-patient as random effects, where time is defined as number of days starting from randomization.
The primary efficacy variable was the annual rate of decline in FVC from baseline to week 52 and was analyzed by random coefficient regression model, assuming linear decline over time.
>>I take it you're referencing the discussion by Neff regarding using a different dosing schedule to get a more efficacious cMin.
Yes - what has made this drug hard to develop is its unusual PK profile. The reasons for this are quite complex. Most proteins present in the body at a low concentration are produced very slowly. But CTGF is different - it is both produced and degraded more quickly. So when the mab attaches to its target, it too can get degraded.
The consequence of this is that some patients (who are producing CTGF at an unusually high rate) will need higher or more frequent doses to achieve the same Cmin because the drug will be removed at a higher rate.
Note the company hasn't discussed this issue at all - the above is my conclusions based on some publications by others.
No I think they will partner (and partner well) based on this trial. Reading between the lines of the CC the drug was underdosed here, and in the Phase III I expect even better results.
>>.no need to raise capital there
Manufacturing for Pam is going to require a massive investment. But maybe that's why they need to partner.
Peter
>>5 out of 997 on 4mg arm during placebo controlled period
That is more concerning. I was just going by their PR which didn't break the events down into subgroups by dose.
Wonder if this is an off-target effect? That was never established one way or another with ponatinib's vascular events.
Peter
>>observed imbalance in thromboembolic events
Out of some 3,100 patients there were apparently five events:
Thromboembolic events – diagnosed as deep venous thrombosis (DVT) and pulmonary embolism (PE) – were reported in five patients receiving baricitinib during the controlled period of two of seven completed Phase 2 or Phase 3 trials in RA. Although an imbalance was observed during the placebo controlled period of the RA clinical trials, the rate of these events in the overall baricitinib clinical program was consistent with that seen among the general population of treated RA patients.
Didn't note much new there. Notably no mention of PD1 combo.
This preclinical paper on SIRPA-inhibited macrophages from a U Penn group is interesting:
http://www.cell.com/current-biology/fulltext/S0960-9822(17)30706-6?_returnURL=http%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0960982217307066%3Fshowall%3Dtrue
>>clearly vendors of the reference drugs are going to have less pricing leverage with payers than they did before FoBs arrived.
Agreed. But a lot depends on just how tough the insurers/PBMs want to play. Their extreme position could be that they would only pay for the biosimilar.
But there is still the issue of docs not wanting to switch patients already on the original drug. If say half of doctors don't want to switch for these patients (and are not forced to by insurance), then a rebate structure based on reaching some threshold still has teeth.
Vinay Prasad is basically a "true believer" that expensive advances in oncology are mostly fake or misleading. For example, he doesn't believe in using NGS to find treatable mutations because there has been no randomized trial demonstrating the benefits.
>Optune
Forgot about them - was thinking more on the drug side. Haven't ever looked at Optune in detail - always seemed like black magic, but I guess they did show some efficacy.
If McCain were to use them that would be very good publicity.
No, there is basically nothing new in GBM. Median survival a bit over 1 year.
>RTRX
There is the minor issue that Shrekli basically stole from Retrophin to pay back these investors:
http://ir.retrophin.com/secfiling.cfm?filingID=1193125-15-292581&CIK=1438533
You don't even get perfect concordance between different parts of the same tumor on biopsy or between tumors and mets.
The key for me is driver mutations, and I suspect that there would be better concordance with those than with incidental mutations. If you find an actionable driver mutation in a liquid biopsy, then that is indeed actionable. So the real question becomes how often those mutations are found in a solid tumor biopsy without showing up in a liquid biopsy.
Trouble is that taking statins is also correlated with fewer deaths from car crashes - it's generally an overall marker for general good habits and good health.
>>M281
Folks can read about the concept here:
https://www.ncbi.nlm.nih.gov/pubmed/24404896
I'd also add that the timing of Soriot leaving doesn't necessarily signal bad news for MYSTIC. By the time he actually exits he'd still reap the benefit (or penalty) from the trial results. And arguably this negotiation must have started well before they had results.
>>BMY/AZN—If you can't say whether a positive MYSTIC result is positive or negative for BMY, then it logically follows that you can't say whether a negative MYSTIC result is positive or negative for BMY. So, it seems odd (but not impossible) that a rumored decision by AZN's CEO to change jobs would cause a sudden sell-off for BMY.
Well of course the issue is not whether JQ can say whether BMY and AZN should be correlated on this binary event, it's what the market currently thinks (whether correctly or not). And currently the market seems to believe that a negative MYSTIC result would also be negative for BMY.
>>"Ronin"), collectively the second largest stockholder of Peregrine Pharmaceuticals
That statement right there undermines any credibility they might have. :)
As I replied on twitter:
I wonder if the authors will publish another article saying "Never mind - turns out that the data was *better* than we thought."
I wonder if the authors will publish another article saying "Never mind - turns out that the data was *better* than we thought."
— Peter Suzman (@Biomaven) July 3, 2017
If I understood the basis for "it could be worse," it was that if the censored patients in the drug group did worse than the censored patients in the control group, then the overall results flatter the drug. But the reverse could equally be true.
There are some trials were censoring caused by withdrawal for tox ahead of progression might skew the results, but I don't see this as one of them.
>>Keytruda in lung cancer may not be as good as we think
Or it could be better...
I wonder if the proxy voting muddled "withhold" and "against" which actually mean slightly different things:
https://www.sec.gov/spotlight/proxymatters/voting_mechanics.shtml#definitions
Always been a bit mysterious to me, because if you hold your shares in a margin account then I was under the impression the broker can freely lend them without your permission and without paying you anything.
Were your shares originally in a cash account?
Peter
So do you think these roxa numbers are also probability weighted? That's what Leerink does and for them you have to back out their 35% probability of failure.
Thanks - would need to know if that was all roxa or included their mab. If roxa royalties, would then need to estimate drug sales - figure 3.5x royalties is about right.
Peter
So does anyone have access to a Credit Suisse analyst's report on Fibrogen and/or AZN? I'm looking for their 2021 or 2022 roxadustat sales estimate.
Trying to correct a very likely erroneous report from Evaluate on the top 20 late-stage compounds.
Thanks,
Peter
>>FGEN Orphan drug designation?
Doesn't carry any great significance - this designation is largely a function of the disease, not the drug.
Yes, I believe this to be correct, and it is the basis for many of FGEN's issued patents.
High hepcidin prevents proper use of iron, and giving IV iron makes things worse. Thus patients with high hepcidin need both very high epo doses and IV iron to achieve even modest gains in hemoglobin.
For these patients with high hepcidin, you can actually reduce roxa dosing over time. With epo you tend to need to increase dosing to maintain the same hemoglobin levels.
This preclinical paper explains why FGEN's roxa is very likely not a cancer-promoting agent:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5354531/
In particular they way the paper distinguishes what roxa does from VHL-driven renal cancer is instructive.
>>smoked cigars
Cigar smoking not particularly associated with lung cancer as typically cigar smokers don't inhale.
These same folks claimed that most e-cigarettes are also harmful in their model.
>Respectfully disagree that MRI is more reliable than biospy for detecting changes in fibrosis.
Fibrogen is trying to push high resolution CT as a measure of fibrosis in IPF. There seems to be decent correlation with clinical measures like FVC.
That's not going to detect relevant mutations for companies like LOXO.
Surveillance using liquid biopsies is generally going to be a nightmare because of the potential for over-diagnosis. The more sensitive the tests the worse the problem will be. But of course for diseases like ovarian or pancreatic cancer early diagnosis could be life-saving.
>>What separates Neogenomics (NEO) from other players such as FMI?
They are kind of playing in different spaces - FMI trying to be value-added, while NEO is a more traditional testing service. Clearly room for both approaches. NEO always dealing with dropping prices though, so it seems like they have to run faster and faster to stay where they are. But hopefully their margins will increase some once they have fully integrated Clarient.
NEO should benefit from increased PD1 testing.
This is a good discussion, but I think the LOXO task is generally much harder than BPMC's, at least in the near term.
For BPMC, all the patients with SM or GIST will get tested - those are going to constitute by far the majority of their potential population.
RET I think at this point is pretty much a given seeing that all NSCLC patients should be tested for ALK and it's a no-brainer to add RET.
Neogenomics might actually be the long-term play here.
>>TRIL
I assume they are paying for the nivo themselves and didn't want to get locked into one drug by having the owner pay (as has been fairly common in other combo trials). I also assume they chose the company who is most likely to want to partner if the results are good.
Be interesting to see what cohorts get expanded - I assume rituxan combo is one of them.