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.
Hi
I am not so worried about people dying from another disease during the trial as this is already factored in our models, and that’s why we talk about “relative survival” VS “overall” or “observed survival”.
All the models compare:
1) The observed survival curve from the figures we know or assume from Cel SCI IT-MATTERS clinical trial (knowing the dates of enrolments – know some events dates – and assuming an end of trial date to reach 298 deaths)
VS
2)The Standard of Care survival curve.
This SOC curve has to be based on “overall survival” statistics to be able to be comparable. We can use for instance data from SEER DB, but most of the time they provide relative survival stats so one needs to adjust them.
The following link is interesting because it gives a hint the relationship between both :
relavive survival
(All phases observed survival is 56% and relative 65%)
The more we wait with no news, the more this analysis looks good as this shows that 1) is much better than 2)
So, my question is about the “evaluable” patients. If you remember well, 2 patients from the 21 had not HNSCC in phase II trial. Also Cel Sci was often mentioning 780 evaluable patients VS 880 enrolee
So if 298 events is related to the 794 patients from the 2 arms we are looking at (The third arm which accounts for 1/7 of the total 928 not being considered for the primary endpoint) the survival is currently really good after all these years : 497 survivors / 794 = 62%, far above any SOC statistic
But if 100 patients out of 794 were disqualified during the course of the trial because “not evaluable” (I am not saying those who died of other cause during the trial), then the 298 observed (not relative) count as counted by Cel Sci relates to 694 and then the survival is 396 / 694 = 57%. Slightly less than before.
I guess that’s another 10B $ question ??
Have a good day
Fosco
Thanks M. Pijoe and congrats for your career too
Hey
I have no doubt they'll be able to raise cash to complete PhIII in any manner ! This $%/@! road has been too long you, wonder how they did to dig millions of $$ under the carpet (well some of them at investors expenses I can tell from experience)
I just hope the existing warrants (eg already anticipated dilution) + cash on hand will be sufficient to complete this before PhIII announcement. This would leave a float of 50M shares approx before the Big Bang !
I would prefer additional dilution to happen after the Big Bang (or the small Bang)
So I'm just sayin' : announcement before August : no additional dilution. After August : probable additional dilution. Now it depends on how pps behaves as some Warrants have a higher price tag.
Hi
For what's worth with Cel Sci previsions (eg always wrong) :
"The Company estimates it will incur additional expenses of approximately $6.4 million for the remainder of the Phase 3 clinical trial" (Starting 1st of Jan 2019)
Cash burn : I would say rather 1,7m than 1,2m per month (5M / quarter). Am I wrong M. CPA ???
6.4 M / 1,7 = 3,7 months + 1/1/19 = end of trial : 23 April 2019 !!!
Now, how long can cel sci operate with cash on hand :
8,7 M / 1,7 = 5 months without warrants : end of May.
May be July - August with existing warrants structure, depends on pps.
More cash will be required if no news before October
Thanks lots Sab
Of course if MK is successful Big Pharmas will have a look at it for H&N but for other types of cancer and indications
This will make the difference between a 20$ stock and a stock @ 200$. That's also why I do believe a possibility of this @200$ figure and, in case of success, I will urge people to wait & see. But at this point in time, let's be realistic, only matters the success of this trial and 0$ is also a possibility
in my 40+ Yo life I made a handful of good investments and 1 very bad one
Very good was to ask my father to buy 30k$ of Biogen in 1994
Good was to buy some tech stocks before the tech bubble
Very good was to create my company with 3 fellows in 2003. (Now we are 100+)
Very bad was a bad timing on Cel SCI which made me lose 100+k$. Could have lost much more if I hadn't sold in lack of confidence
Very good was to buy real estate in South Coast in 2013 and rent it.
Very good was to buy HGSI, right before Benlysta was proven effective
On HGSI, I remember drug was proven ineffective on a first trial (so the stock price was very bad), so they took a subset of patients and rerun a trial. I had a good feeling when the CEO was telling "it has to work it has to work". He was right it worked and 4 to 40 $ in the following days !
Now Geert says "it has to work it has to work". I would like to believe him but I have a good feeling. Still licking my wounds though from the BS of the past
I have been listening to the webcast, sorry hadn't had time before
I do find the tone convincing
1) He mentioned cancer stage III AND IV. That's an important point in our models as I have used basically stage III OAS curves, because IV includes metastatic which is not in the inclusion criteria.
He keeps on speaking on 55% survival at year 3. He reiterated that by the time the trial has been lasting the drug must be effective.
If he is right about the SOC survival, there is absolutely no doubt that the drug is very effective.
2) Reiterated that he didn't have access to the data but that IDMC and FDA had. And therefore allowed trial to continue in August 2018 having seen it.
3) Did not mention a date, but said 2019 will be an interesting year.
So my conclusion for it remains the same. The question is whether it is effective or very effective. If SOC is really nearly as bad as he mentions, no doubt it is very effective and endpoint should be reached before mid of year. Then he will need a bit of time for analysis and cleaning all the data, but then expect a result by Q3 or Q4 and a very very interesting year 2019 !
Kpauliukonis
I don't know what is blinded but I can rely on trial registration document and it says clearly that it is un-blinded, and the reason you mention are absolutely valid (it couldn't have been blinded) which means that M. Smith who is taking the drug is recorded in the DB as having taking the drug and M. Vangala will be recorded as member of the SOC group. Patient is un-blinded and doctor is un-blinded.
So the data itself is recorded with this information. Whether it is for the CRO eyes only or encrypted until 298th event, you have to rely on management word, not on facts. What you don't know is if they really don't know or if they just can't reveal the figure. It's easier for them to say "we don't know" than to say "We know but cannot disclose because of the risk jeopardising the whole study" (it ain't over till it's over).
It's exactly the same for the number of events. Nobody thought about asking them ? If so, what was the answer ? We know for a fact that they are monitoring this figure. We know they know. They know also that we know that they know. That's why may-be they removed the planned end-date from early 2019
Thanks Sushi
So what does this mean ? Simply that what we suspected based on stats was right : trial will last longer than initially announced. That’ good. Just hoping they don’t take too much time between 298th event and data publication.
lightrock
you ask a good one
That's the difference between "relative survival" and "observed survival"
Relative is observed divided by normal survival of a similar population (gender, age, race). While I am still looking for reliable sources of data, you can have a hint there in fig 6. Normal survival = observed/relative
link :
https://www.ahns.info/resources/oral-cavity-cancer/4/relative /
Sushi,
No credit at all, the link to Taïwan hospital was provided by a poster from Stockwits
This said, more useful than speaking with an analyst, I would like Cel SCI to explain why they do not report events. They know the numbers. We, as investors deserve to know those numbers as well. This is fundamental information needed to follow the trial and there is no reason why a small group of people would be allowed to know it and not all those who have put faith in the company.
I am reacheable @ fosco.sinibaldi at yahoo.fr
Fosco
Btw : what a nice white tennis suit you bear !
how did you know I am fan of MWI, everett and deutsch
I can't believe it has been proven 100% wrong, as it matches so well QM. I hate copenhagen interpretation.
Although my own interpration is a bit less extreme. I believe each observer follows a specific path. Consciousness is key. Backwards causality is key also to understand how present is built from the future (final condition being at least as important as initial condition). There is no such thing as determinism, but may be a two way determinism. QM can be derived from simple interactions happening both ways of the time arrow. Your cat could be very well dead for you but still he would feel alive in his own universe/world ... lots to say about it, lots of ideas !
Take care
Fosco
Sushi
Zacks had it wrong on H&N cancer survival. They can't have it all right, right ?
Geert initially indicated "early 2019"
We managed to get survival figure from feb 2017 and feb 2018 which indicated 133 and 208 events. This alone sets an earliest possible date for April 2019. More sophisticated models provide October. Strong uncertainty do persist on the date but it won't be february. It could be earlier (july, august,...). You're a CPA, you can do the math yourself
After we got this finding from a member of the MB, we communicated on it and its consequences, the link, from a Taïwanese hospital was... broken
Geert didn't calculate right and voluntarily concealed those figures for specific reasons.
I can guess some reasons, I'll let you provide others if you want:
- October 2019 would mean refunding (dilution) coming while earlier date would allow earlier speculation. We now have heard about those pending warrants.
- Geert personnal revenge against those who put obstacle through his way. He knows he is certain on of success. He can play with the investing community
-May be this means bad news, but I can't see how
Fosco
wow you impress me Lightrock, you are true physicist !
ain't no physicist, but got my own predictive tools.
My tools for instance are saying :
98% probability your first name starts with an A
85% you live in Austin, TX
74% your wife first name starts with K
51% we win big with CVM
what are you doing ? Calculating Pi ? That's all I can do with Monte Carlo !
Alexander,
I agree with most of what you said , but not all of it.
Try to follow the reasoning below, you will see, it is pretty simple, just a bit of concentration is needed :
I think you omitted one major point. We know for sure two events dates : Feb 2017 : 133 events and Feb 2018 : 208 events. This tells a lot and allows to calibrate our statistical models as Lightrock is trying to do. We are not doing blind statistics !
So :
133 deaths in feb 2017 that's 661 survivors
208 deaths (+75) in feb 2018 that's 586 survivors, yearly death rate was 11,4% (75/661)
As we know death rates diminish in time (that's a from cancer epidemiology), lets apply 11% death rate to the 586. This means 64 died since last year (11%x586). So today's count should be around 272, it can't be much more.
This allows to compute an end of trial date (we still need 26 deaths to reach 298), let's say 1st October 2019
With all this you have 4 points in time : feb 2017, feb 2018, feb 2019, Oct 2019. You can draw a survival curve which will be pretty accurate.
That's all I did, with precise recruitment numbers, logarthmitic curves, assumptions on MK benefit etc etc....
So what it shows, is that the five year observed survival from the WHOLE (TEST + MK) group should be close from 60%. And this is what we want to know, because in whole Head and Neck cancer survival DB we find SOC survival curves much below. For instance, if you take the US benchmark, through the SEER public db you will find 2008 - 2014 five years 65% relative survival for regional Oral Cavity cancer which implies a 56% observed survival (relative is observed less other causes of death). And the US benchmark is on the best in the world. And that's where I agree with you, germany will show something like 50% observed survival (check litterature), UK 56% in best of cases, and India under 50%. Most of the study took place in countries like india, ukraine, hungary, etc... where SOC 5 year survival is likely below 50% for stage III-IVa H&NSCCC
Blinded study ? True, but 60% five year survival in TEST + CONTROL, that means thas SOMETHING is likely to be happening in the TEST arm improving significantly the survival of the whole group of 794 patients.
Funds do have tools to follow this, much powerful than we have. Blackrock is not Lightrock (no pun intended) or Fosco. They probably know all we know and know they have time to invest and to minimize risk they can gradually invest until october. I believe will see more money pouring in as key threshold months are left behind.
Is it more clear Alexander ?
Wow
This is getting too complex for me to follow, sorry I don't have really time by now
I don't know much what "30 events" means, what I do know :
When endpoint is reached, a 10% improvement would mean :
161 dead in SOC
137 dead in MK
236 alive in SOC arm
260 alive in MK arm
Difference is 24. Wider the gap, the better
My Tab 4 in my sheet matches , Feb 17 and Feb 18
Today count should be 271. Endpoint should be reached by 10/15/19 What is stunning is that I didn't have to change my SOC survival curve, I only have to tune the MK benefit to 25,2% to make the event log match those 2 dates almost perfectly
Year Survival%
1 84
2 71
3 63
4 57
5 54
6 49
7 46
8 43
y = aln(x) + b
a = -19,45
b= 84,273
MK VS SOC : + 25,20%
A SOC that would match the same pattern (the two dates plus endpoint in H2 2019) with 0% benefit in MK group would be something like this :
Year SOC
1 94,9
2 79,7
3 70,8
4 64,5
5 59,6
6 55,7
7 52,3
8 49,4
9 46,8
10 44,5
y = aln(x) + b
a = -21,91
b= 94,891
food for thoughts. I would love to have up to date SOC recent (2010 till 2018) data for stages III + IVa HNSCC
Trustbaby,
Well it depends on SOC survival, the big unknown, the 10B$ question.
My model for stage III+IVa cancer, with 84% Y1, 63% Y3, 54% survival at Y5, says we passed the threshold in January (which means that should the 298th even have occurred in Jan a 10% benefit would have been noted) and should reach primary outcome (298 deaths) by October with 25% benefit in MK arm
Dear Sushi
I haven't changed a mere bit in my sheet
I was discussing Lightrock sheet, as I am interested in other angles of study in particular his computation of response rate. I did find some strange things and I did question them.
I am not sure I get still all the substance of it but I think we are approaching.
Still, I am not sure what Lightrock means by 3years OAS as primary endpoint of the study because, as, per CT it is written :
-Primary Outcome Measures :
Overall Survival (OS) in LI + CIZ + SOC vs. SOC [ Time Frame: 3 year ]
>>> Time frame here is an estimate of when data is unblinded : 3 years after Actual Study Completion Date (Dec 2017).
When 298 deaths occur, they build two Kaplan-Meier survival curves (with 298 points on the graph), compare observed survival on one arm VS the other. The logrank test allows to validate statistical significance (need p<0.05) of counted survival benefit regardless of time distribution of events in each group. Both of curves will show OAS from Year zero to Year 8 if it shows a recruited patient in 2011 who survived 8 years.
yup
easier to count people than % : for instance if soc arm survival shows 17p alive, MK should show 18,7
Now I can see you are changing this on gsheet . Beware your formula $L13+1.1 as it should be $L13*1.1
Hi again,
Impressive work !
But sorry to bother again
As my own sheet did not match
I noticed that, in your columns "If 100% of the TEST GROUP is 10%
better than control group" actually you added 10% to survival %
For instance in AB13 your formula is =... (K13+10)/100...
Well a 10% improvement is not this. For instance if survival in K13 is 50%, this will give 60% and therefore a 20% improvement VS SOC. Is this intended ? If not, the right formula should be ... (K13*1.1)/100...
(This said I prefer to keep improvement as a variable to calibrate with the dates)
Take care
Fosco
Hi
See a quick and dirty test on your column L
https://docs.google.com/spreadsheets/d/1oe89En9CYKDX8z3kaExsIbSsAgNvZ5nL9k98-1ZTy3I/edit?usp=sharing
XL computes the regression based on 3 points (from by UK cancer data) : Initially b was 78,573 but I changed it to 78 to match your 55% at Y5 in order to measure the impact on changing your straight line survival curve by a log curve. See how the log curve matches accurately the real OS figures at Y1,Y5,Y10
Initial regression a= -14.32; b= 78,573
Year UK H&NCancer DB Survival% / REG LOG%
1 78.4 / 78.6
2 68.6
3 62.8
4 58.7
5 56.1 / 55.5
6 52.9
7 50.7
8 48.8
9 47.1
10 45.2 / 45.6
you can have 90 at Y1, what counts is :
1) the initial slope
2) the flatnessness afterwards. All OAS curves shown so far on H&NC have shown flatnessness
I remade(quickly) your computation with your V2 (55%) data and a log curve (78% at Y1) that mimics UK data. I found really a great gap with your initial analysis
y = aln(t) + b
a = -14.32
b= 78
A good guide on survival curves here :
https://www.cancerguide.org/scurve_basic.html
Excerpt :
"Many curves show a decreasing risk of death over time (If the risk decreases to zero you get a plateau!). Some curves appear to have a roughly constant risk of death over time. A constant risk of death leads to an curve which represents exponential decay like the decay of a radioactive element. In a perfect mathematical world, an exponential decay curve doesn’t actually reach zero, but with any limited number of people to start out with, such a curve will eventually reach zero. It can be very difficult to detect decreasing risk by eye – but the evidence is that even for advanced cancers with a poor prognosis there is a good chance the survival curve is decreasing risk"
Lightrock
I had a look at your renewed sheet. I believe there is a fundamental issue in it which lies in cells H5,H6,H7. Death rates are not linear. They are high in the first years and decrease in time : survivors are those getting cancer resistant. The curve is getting flat.
You can't even say "2 people will die every Xmonths". You could eventually say 2% of survivors die every X months. That's called a constant death rate "k" in time (dS=-kSdt) and this would show an exponential curve (like in nuclear decay) and a not linear curve. But even the rate is not constant, it diminishes with time.
The more appropriate approximation seems to b e the log curve which is getting flat in time. If you take a look at UK survival data (where we have 3 points in time : y1, y5 and y10), it matches quite well the three point.
So the impact in your, otherwise remarkable, work : you over-estimate survival in the early years before Y5 (which were most of the patients come from) : therefore your survival is too strong with SOC, or you have to show a low rate of responders (24%).
I cannot conceve, if survival in MK group is good at data readout, that they would wait 6 months before publishing anything... This would be bad for shareholders, bad for the stock and would prevent easy funding. Besides concealing it would be best way to leave it to unauthorized leaks and insider trading as wider audience would be in the know of unblinded data.
Indeed they would cautiously need to ask their lawyers what can be published because of SEC, FDA, signed NDAs, uncertainety on data (are reported deaths real deaths ?) etc,... but still this would be the kind of secret not to keep to oneself
"Understood this is conservative in sense that SOC likely has not improved this much"
... fact is that we won't know before the study comes out as study is taking place in 100+ sites worldwide, obviously SOC is not as good in some countries than in the US. I have seen very bad 5 year OS in some countries, and better in others (like the 56% in the UK study).
Here I have taken 53% which is better than the 43% OS mentionned by Cel SCi in the CT db
Dear Georges
Let's be precise in the assumptions
1) Based on the following conservative SOC Overal Survival for cancer stage III and IVa (cel sci inclusion criteria) :
Year Survival%
1 84
2 71
3 63
4 57
5 54
6 49
7 46
8 43
2) Having calibrated the events table spreadsheet with % MK improvement benefits so that the 2 dates where we know the number of events do match the model (Feb 2017 : 133 events and Feb 2018 : 208 events)
3) I can output an end point being reached by October 2019, and MK benefit being aroung 25%, which matches Cel SCI satements that study will complete by 2019 (rather second half than first half)
THIS SAID :
==========
* Due your own DD and make your investment decision accordingly : I can be proven wrong by any SOC survival proven really much better than anything existing so far in litterature for cancer stage III and IVa
* For those who don't believe in mass predictions made on invididuals behaviour which differ from individual to individual, I will answer by this quote from L. Boltzmann : "Nothing is more practical than a good theory". (Boltzmann is the physisics who discovered the kinetic theory of gases and who was able to predict macroscopic behaviour of gases (PVT) based on the only fact that they are made on invidual molecules behaving randomly)
sorry duplicate message
doingmybest
Many reasons to doubt:
- these figures are from early 2000
- geert stated himself that death were too slow (in 2016)
- many studies have emerged since :
https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/head-and-neck-cancers/survival#heading-Zero
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0053415
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3548847/
https://www.ahns.info/resources/oral-cavity-cancer/4/
http://www.cancer.ca/fr-ca/cancer-information/cancer-type/oral/prognosis-and-survival/survival-statistics/?region=on
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5384671/
https://www.sciencedirect.com/science/article/pii/S0360301618307910
https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/head-and-neck-cancers/survival#heading-Two
https://www.slideshare.net/doctorbobm/head-and-neck-video-1
https://seekingalpha.com/article/4170868-cel-sci-simple-math-appears-imply-phase-3-success?page=2
doingmy best
I am reposting for you my spreadsheet hereunder
As lightrock said, you need precise numbers because the mere one event in one arm or the other matters a lot
CVM initial soc survival model : 55% Y3, 43% Y5 was way too pessimistic (tab1). I assumed a more conservative approach on tab2 (63% Y3 / 54% Y5). This is for the inclusion criteria of the study : stage III and IVa (non metastatic) H&N cancers.
Tab 4 is same model as Tab2 with events dates matching real events as measured in feb 2017 and 2018 (and better survival : end date oct 2019)
https://docs.google.com/spreadsheets/d/1qW36sMXgKvx2_76GPF3fjmXFYApgsOUq0vU1qlumz6w/edit#gid=1323197717
I agree 100% with The_Q.
Feurstein is not credible this time, the Johnny in him has too much taken on him or reversely But nice try !
Regarding S/P in the eventuality of a positive outcome, my recommandation would be not sell at any price until a buy-out happens _and it will happen_ because it could be anywhere between 1B and 10 B ! 1B would be the minimum I believe (20$ ps)
He didn't sell any
This is an update (re-registration) of an old S1. See June S1, Geert has same number of warrants.
When (and if) an insider (buys or) sells, we'll know through the SEC Form 4.
Fosco
CIZ is in test group only (CIZ + LI), SOC arm and third arm don't have CIZ. Cyclophosphamide : this chemo is not boosting the immune system, on the contrary it inhibites cells replication. Tumor reduction has been noticed with Cyclophosphamide, but not as much as in this ph II trial with LI (42% response rate 30% to total tumor reduction). Thus the idea of the combination is to first paralize the enemy (injection of Cyclophosphamide 3 days prior LI) and then attack him by boosting defence cells with LI.
Ain't no oncologist but that's my understand for what's worth for
Take care
Fosco
staticmirror79
What you hilighted is key to understand why cel sci is considered as a risky "stock". Because little improvement is required to get it marketed (eg 10% better survival) it can only be proven statistically in a large population, eg only be proven in phase III.
Phase II was very limited but that's how all phase II work
Phase II had no placebo arm. True but how can you compare a 10% improvement on a very small population. Basically in phase II you could only do a qualitative assessment : total responders, reduction in tumor size, etc.. and say that's what we wanna see in a larger scale.
That's what Cel SCI did and with such challenge to prove that the drug works as pre-treatment to SOC there were little other ways to conduct a phase II.
So when Feurstein said "phase II don't prove anything", he is right but like in most of phase II, that's why there is a phase III for drug marketing approval !
Williams
At last some words of reason amongst an ocean of dis-information !
The only new information in this prospectus is the creation of 500 000 shares to pay Ergomed. As it says it is an update of existing prospectus that has expired.
The rest of the information is the disclosure of who owns warrants and have declared that they will exercise it one day between now and 2022, which will create new shares, regardless of when they decide to exercise, which is their own business and decision.
For instance : Geert has the same number of exercisable MM warrants (147 929) from past June S1. He didn't actionate them back in 2018, why would do it now ?
lightrock
I came to the same conclusion a while ago
The "futile" came from this unblinded data, but the data set was way way too small for a descent statistics and the time elapsed way to short to draw conclusions on events-based survival. Therefore I think they worried more about safety than efficacy.. Or they had a double agenda paid by nasty big pharma...
Spidey,
I try to figure out why there is so much differences but I am not quite sure why there is so much as I used your SOC data, may be I made a mistake somewhere ?
My model is using an extrapolated log curve. You have to agree on the numbers on this log curve that they match your figures (if you have only 62% OAS at year 5 it will be hard to extrapolate, you need more points in time like 1 year, 3 years ... the more the better). Why log curve and not linear ? Because survival works like this ! Means survival dynamics is strong in the first years and getting flat after (which means basically that the weak die first and when they are all gone the strongs remain). I noticed you using a linear model for events ("23" every 3 months) to compute events per date, therefore I believe you don't take into account the flatness of the curve and this might explain the difference. In any case I agree with your overall statement.
Source for Feb 2017 is the chinese document : 75 events in the reporting period from Report No. 8 from CVM
"1) The company entrusted the director of the hospital's otolaryngology head and neck, Zhongyu Wang (???), to conduct {an open, randomized, multiple centres clinical trial to compare the use of Multikine (Leukocyte Interleukin, Injection) plus standard therapy (surgery + radiation therapy or surgery + radiation combined with chemotherapy) and the use of standard treatment in primary advanced squamous cell carcinoma of oral and soft palate cancer (program number: CS001P3; IRB number: SF14218).} By reviewing the registered clinical laboratory result, the case is listed on August 25, 2014, Zhong Rong Ren Zi Zi No. 1030020068 was approved.
2) The safety of the drug Multikine has not changed since the seventh safety review of the drug development report. A total of 208 deaths have been accumulated so far. During the reporting period, there were 75 deaths at home and abroad. This trial was suspended from September 26, 2016 and has been notified by the Institute for reference. It was released on August 10, 2017 during this report period. In addition, there were no new safety or efficacy issues related to drug treatment, and no SUSAR was involved.
3) A copy of the Multikine Drug Safety Development Report (Development Safety Update Report No. 8, Date of Report: 16 Mar 2018, for the reporting period from February 4, 2017 to February 3, 2018) is sent to you for reference."
Sushi,
You're the detail guy, I appreciate,
My latest conclusions and now almost convictions are that
- MK is efficient (Overall at least 25%)
- Endpoint will be reach later than we expected (we expected Q1 2019)
Your sense of details confirmed them
There can't be other way, with 208 events in Feb 2018 (and 133 in feb 2017) you need more than one year to get 90 more events... I would say that you need 15 to 18 months
Cheers
Fosco
Lightrock
It took time to me to assess what you meant by "response rate", I wanted to make sure of what you meant.
Now i understand fully your point, I didn't compute this in the model, because models being models you can be so wrong with the first assumptions that needless focus on the details. Indeed it does shorten endpoint timeline, "the cherry on the cake"
Take care
Fosco