Scancell Share Discussion Threads

[ruckrover] 25/4/2024 09:40

Inan,Me: Any more red herrings you want to throw in?You: YesAt least you admit it LOL! Hope you are ok?

[inanaco] 25/4/2024 08:11

tedious

but this is your Nemesis Ruck ................ its yourself

------------- ""My job involves working with probabilities.""

[ruckrover] 25/4/2024 08:05

Try and understand the concept and discuss it rather than nit pick.It's you that can't divide 18 by 26!

[ruckrover] 25/4/2024 08:03

I used 11 out of 13. 7 plus 11 is 18

[inanaco] 25/4/2024 07:39

but this is your Nemesis Ruck ................ its yourself

------------- ""My job involves working with probabilities.""

RuckRover

Posted in: SCLP

Posts: 5,369

Price: 9.60

No Opinion

RE: Onclive23 Apr 2024 10:21
Bermuda, Moonparty,

Many thanks for the response and clarification.

Even so, regardless of whether success is 85% or 70%, where does the 90% confidence come from?

My job involves working with probabilities. To achieve a 90% confidence level (assuming a normal distribution), you would need 1.28 standard deviations. But deviation from what. I guess if we had the %responses for each individual patient in the 13 measured to date, we could work out the standard deviation and apply 1.28 times this to the arithmetic mean. We could then classify the results and work out the numbers for complete response and partial response and see if this is over 70%.

But I don't think we are privy to enough data to see how 90% confidence was arrived at.

[inanaco] 25/4/2024 07:22

start again ruck you could not add up, 7 plus 9 = 18

LOL

best of luck with your investment try wet wipes on your brown nose

[ruckrover] 25/4/2024 07:14

"you would not fail the trial at 69.2% ruck"Perhaps so but maths is precise and 69.2 is LESS than 70.How far do you want to move the goal posts?How many more red herrings are you going to throw in to mask your errors?When are you going to get your calculator fixed?When are you going to admit you don't know more about the science and probability than professor Durrant?

[inanaco] 25/4/2024 00:03

RuckRover24 Apr '24 - 23:55 - 7968 of 7971
0 0 0
Also, we are not talking about phase 3 trials. LD was talking about a 90% chance of these results being replicated in a larger group in the CURRENT trial.

Any more red herrings you want to throw in?

-----------------------------------------------------------------------------------

yes but after each trial you get a new probability .......... exactly what i have said

but if efficacy is the same at 85 % the probability in a bigger trial will be higher than 85%

simple ...

dont try and twist it ... to many posts on here to confirm what I have actually said not the twisted version Bermuda uses

[ruckrover] 24/4/2024 23:57

Catch you in the morning- time for bed

[inanaco] 24/4/2024 23:56

you would not fail the trial at 69.2% ruck

[ruckrover] 24/4/2024 23:56

Oh, you already have!

[ruckrover] 24/4/2024 23:55

Also, we are not talking about phase 3 trials. LD was talking about a 90% chance of these results being replicated in a larger group in the CURRENT trial.Any more red herrings you want to throw in?

[inanaco] 24/4/2024 23:54

MHRA and the FDA use probability to approve

they may test on 1000 patients ............ but the real world test maybe 100,000/year or more patients

should we trust them ? if you cannot predict or scale up the trials to real world numbers

because Bermuda said you can't

[ruckrover] 24/4/2024 23:53

Nobody's denying that the larger the numbers the more accurate the predictions will be. That's universally accepted but we aren't there yet.

[ruckrover] 24/4/2024 23:51

I think we've had this discussion in the past - you need to get your calculator fixed.18 out of 26 is 69%Come back when you've learned simple arithmetic.

[inanaco] 24/4/2024 23:44

more numbers the more accurate it becomes ...

fully aware of that ...

but i am NOT predicting the phase 3 trial on 13 patients

but potentially 80 plus

but the maths would not change .... if the efficacy was 85%

the probability would still be higher than the efficacy because you only need 70% for a pass

[inanaco] 24/4/2024 23:39

""So let’s double the people on the trial to 26. In the second group of 13 there are 7 people with a response. This makes a total of 18 responders out of 26 which is below 70%"" Ruck

18 of 26 is 70.2%

success



scancell 9 + 7 = 16 or 61%

failed

[ruckrover] 24/4/2024 23:29

Inan,"has to be higher than 85% because you have a 15% margin of failure and still hit your pass markcommon sense mate"It's far from common sense!So let's double the people on the trial to 26. In the second group of 13 there are 7 people with a response. This makes a total of 18 responders out of 26 which is below 70%So in order to be above 70% you would need at least 8 responders from the second group of 13. So Scancell are telling us that there is an above 90% chance of there being 8 or more. How do they work this out? I don't know. No one knows if another 13 people would match 11 as in the first group or would be more or less. It could be anything between 0 and 13.

[inanaco] 24/4/2024 23:12

by the way that would be the same trial Biontech is running ... check point failures

they are using their tumour associated vaccine in CPI failed patients

[inanaco] 24/4/2024 23:05

a crossover trial works superb because both arms use the same standard of care treatment

so you cant ascribe efficacy to another treatment like chemo etc as still having an effect on a patient crossed over to receive the vaccine as well because they are on the same standard of care treatment

but now you are treating checkpoint failures ....

Pure unbiased efficacy which cannot be refuted if they respond to the vaccine

[inanaco] 24/4/2024 22:56

what Bermuda is saying it gets more complicated because he is trying to second guess the control arm .... why i have no idea !! that's just as is ... standard of care treatment approved and we already know what the result should look like ... 55%

but that has nothing to do with calculating probability of efficacy in the vaccine arm

""" lol- so in a randomised combination study you're going to ignore the control arm and give the treatment arm a 90% chance of success. How are you measuring success then? It doesn't matter whether it's 200, 400 or 2000 patients the point stands. You CANNOT take the 90% probability of success for the SCIB1 trial and apply that to the next trial. That 90% figure was simply the chance of an 85% response rate in 13 patients turning into a 70% response rate in 43 in that particular trial.""" Bermuda

but if you agree with Bermuda ... OMG !! elementary maths, why would you want to know if the control arm has a probability ? and how does the control arm affect the efficacy of the vaccine arm it cannot

its there to prevent bias not prove the checkpoints ... even if that control arm had efficacy of 60% it would not change the target needed of 70% in the vaccine arm if that is the agreed endpoint

[inanaco] 24/4/2024 22:42

but even so its very simple ....

if your trial is hitting 85%

the probability of repeating that trial to achieve 70%

has to be higher than 85% because you have a 15% margin of failure and still hit your pass mark

common sense mate

[ruckrover] 24/4/2024 22:33

"Claim you are correct then disappear "I'm not claiming anything. I'm saying I don't understand how the 90% is calculated.I'm not saying it's wrong.If someone has convinced LD that's what it is, that's good enough for me.

[inanaco] 24/4/2024 22:07

what i do like Ruck you admit to having NO IDEA ....

then you become an expert ...

are you really this thick ?

RuckRover

Posted in: SCLP

Posts: 5,369

Price: 9.60

No Opinion

RE: Onclive23 Apr 2024 08:55
Violin,

Yes, that would make sense. It says 27 enrolled and 24 dosed. So 3 enrolled but not dosed.

I must admit, I struggled to make some of the numbers add up.

The other puzzle for me is the assertion that there is a 90% chance that the 85% ORR in the small sample of 11/13 will be repeated in the full 43. Whilst there is grounds for optimism there is no guarantee that these results will be repeated. If the 85% IS repeated in the 24/27, then the chances of repeating in the 43 will improve. But how the 90% figure is arrived at I have no idea.