Roundtable Discussion; The Future of Mineral Sands. Watch the video here.
London South East prides itself on its community spirit, and in order to keep the chat section problem free, we ask all members to follow these simple rules. In these rules, we refer to ourselves as "we", "us", "our". The user of the website is referred to as "you" and "your".
By posting on our share chat boards you are agreeing to the following:
The IP address of all posts is recorded to aid in enforcing these conditions. As a user you agree to any information you have entered being stored in a database. You agree that we have the right to remove, edit, move or close any topic or board at any time should we see fit. You agree that we have the right to remove any post without notice. You agree that we have the right to suspend your account without notice.
Please note some users may not behave properly and may post content that is misleading, untrue or offensive.
It is not possible for us to fully monitor all content all of the time but where we have actually received notice of any content that is potentially misleading, untrue, offensive, unlawful, infringes third party rights or is potentially in breach of these terms and conditions, then we will review such content, decide whether to remove it from this website and act accordingly.
Premium Members are members that have a premium subscription with London South East. You can subscribe here.
London South East does not endorse such members, and posts should not be construed as advice and represent the opinions of the authors, not those of London South East Ltd, or its affiliates.
Noted. I will ensure I spell out future comments very slowly for you.
Meanwhile, please feel free to pick on somebody else for your tedious nitpicking.
Really? How can this sentence be read in any other way?
'If there was a 90% chance of exceeding an 85% ORR then you can be certain that 85% would have been the target ORR, not 70%.'
Perhaps 'without saying as much' part could be the problem?
They wouldn’t.
I was just making the point that the 90% confidence in reaching 70% is driven by the 85% already having been achieved in one cohort - and noting (without saying as much) that it would be nuts to think they had a 90% chance of beating 85%.
You seem determined to get the wrong end of the stick today……..
Ee
The whole trial was designed and powered to produce a 70% response rate as that is the level which demonstrates that SCIB1 is having an impact (ie. 20% higher than CPIs alone) and merits further development. How/why could/would they change that to 85%?
Johnny,
It is the fact that 85% is “in the bag” from the initial cohort that drives the 90% probability of exceeding 70% ORR.
If there was a 90% chance of exceeding an 85% ORR then you can be certain that 85% would have been the target ORR, not 70%.
That said, as I pointed out earlier, the 90% number is irrelevant once the results start coming in…..indeed their internal expectations may already be different?
Bermuda,
Thank you for getting clarification on that, I appreciate it.
Johnny,
Yes I know and it was your post (thanks) that prompted me to email as this seemed to contradict the original RNS which I understood to read 70% - see my post of 23rd. I think the wording of that part of the AGM presentation was unintentionally slightly ambiguous and could easily be interpreted as the probability of replicating the 85% response rate whereas they actually meant the probability of successfully passing the 70% threshold as the first cohort had done.
Anyway at least we know now. It will be really good news if they hit that 70% response rate which will mean the trial has been successful and a bonus if the response rate is higher.
Hi Bermuda (and emptyend),
I don't doubt what you are saying but that contradicts what Lindy said at the AGM and what it states on page 10 of the AGM presentation (See my 12.27 0n 23 April).
When you say "Lindy has clarified ............... ", have you emailed her?
Yes that is correct.
And, as someone pointed out earlier. the 90% probability figure is meaningless once the actual results start coming in. The actual results may be better or worse than implied by the high confidence figure.
Johnny,
Just an update on the 90% probability of success discussion regarding SCIB1.
Lindy has clarified that it is the probability of achieving a 70% response rate, not replicating the current 85%. To be precise it's the probability of achieving a '70% ORR on 43 patients' from the current SCIB1 study.
Hi Ciaskin,
I agree with your maths, it would need to be 28 out of 34 to maintain the 82% level. Scancell do occasionally give out conflicting information and I sometimes find it confusing.
I notice that page 9 of the interim results presentation in January’24 fits in with what you say as it makes reference to “more than 27”, it states:-
Simon stage 1 >8/15 ORR; Simon stage 2 >27/43 ORR
https://www.scancell.co.uk/Data/Sites/1/media/fininfo/2024/scancell-interim-results-fy-2024.pdf
Bajookajr’s link states 27 of 43 patients now enrolled, so agree another 7 patients would take this to 34.
I am no expert but I interpret (on the basis that when they got 9/11 ORR in Simon stage 1 they immediately moved to Simon stage 2, rather than waiting for results on 15 patients) that if they get 28 out of 34 that would be enough and they would not require results from 43 patients. AIMO
Thanks for reposting Johnny. What is slightly confusing for me is that while 9 out of 11 is indeed 82%, 27 out of 34 is only 79%. It would have to be 28 out of 34 to maintain the 82% level. Either way, is she saying that if we were to achieve the necessary ORR after just 34 patients dosed, then we wouldn't need any more before moving to/preparing a Phase II? If so, we would only need to recruit another 7 patients...
Many thanks Ray, JonhnnyB1, bazookajr those comments are really useful.
I'll have to work out how to use the binom.dist function in Excel.
ATB
If we want to find the probability of achieving at least a 70% response rate in a group of 43 patients, assuming our best estimate of the response rate is 82% (i.e. 9/11 patients had a response).
We model the number of responses X in 43 patients as X~Binomial(43, 0.82)
We then calculate the probability of having ORR > 70% i.e. P(X>=30), since 30 out of 43 is approximately 70%
If we plug in the numbers, the calculation shows that there is approximately a 96.2% probability of achieving at least a 70% response rate in a cohort of 43 patients, based on the initial success rate of 82% observed in the first 11 patients.
The 90% refers to the probability of repeating existing data. Page 10 of the 2023 AGM presentation states: “Greater than 90% probability of replicating this data in the full cohort of 43 patients .”
https://www.scancell.co.uk/Data/Sites/1/media/docspres/agm-presentation-november-2023.pdf
In the 2023 AGM presentation, Lindy says:-
“Again if you like your stats and I’m not going to argue stats with you because I’m not an expert, they tell me if I get 9 out of 11 responses there’s a 90% probability that we will meet that milestone of 27 out of 34 which is really exciting”.
https://www.youtube.com/watch?v=say9mRBFzCo
This is a fuller version of what Lindy says from about 6 minutes in, talking about metastatic melanoma:-
“Metastatic melanoma until very recently, probably five years ago was a death sentence, there was no therapies doesn’t respond to chemo, nothing works in metastatic melanoma. The double checkpoints have now improved that to 50% our intention was to improve that 50% t0 70% because the clinicians told us that would be a meaningful impact for these patients, but a really tough ask, there have been many combination studies in this area that failed to achieve that but we’ve got better than 70% as you can see we smashed it at 82%. So we were supposed to see 9 patients out of 15, we actually saw those 9 responding patients and I’ll show you that data, out of 11 patients. That’s an 82% response, even more exciting there was no toxicity associated with the melanoma vaccine. The checkpoints themselves particularly one of them can be quite toxic but the addition of the vaccine had no additional there at all which again is really good in combinations. So we now move to the second stage, so the first one we had to see 9 in 15, the second stage we need top see 27 responses in 43 patients, or if we see this 82% response rate it will be 27 in 34 patients. Today as we talked to you as its changed again there it’s 21 patients that have been immunized so we’re pretty much going down that route, we only have to get 34 so hopefully proably early next yearwe will finish that but you need to bear in mind they do need to be on the study for at least 3 months to that first scan when we get that first response to see whether they’ve got an objective response. So hopefully sometime in the next half next year we should get that data and see how good it is. Again if you like your stats and I’m not going to argue stats with you because I’m not an expert, they tell me if I get 9 out of 11 responses there’s a 90% probability that we will meet that milestone of 27 out of 34 which is really exciting”.
Ruck
"But I don't think we are privy to enough data to see how 90% confidence was arrived at."
I would guess that Lindy asked someone from the Maths department at Nottingham Uni about the probability of success.
She herself stated that she was informed of the probability but she didn't know how the figure was calculated.
But I do think that SCIB1's chance of success is higher that Burnley's chance of avoiding relegation 😒
Ruckrover,
“ Yes, that would make sense. It says 27 enrolled and 24 dosed. So 3 enrolled but not dosed”
Not necessarily. iSCIB1+?
Definitely a 99% probability that 99% of posters have 0% interest where it came from.
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.
One live Chanel
22nd of April 24 - ( Linked to relevant content )
Vishal Patel, MD, FAAD, FACMS, associate professor, Dermatology, George Washington (GW) School of Medicine & Health Sciences; director, Cutaneous Oncology Program, GW Cancer Center, discusses the evolving use of PD-1 inhibitors for patients with cutaneous squamous cell carcinoma (CSCC) in the neoadjuvant and adjuvant setting.
https://www.onclive.com/view/dr-patel-on-the-evolving-use-of-neoadjuvant-pd-1-inhibitors-in-cscc
Just rambling now, but this could be a normal distribution (with response rates) or a binomial (with binary response / no response). I used to understand this stuff when I was at Uni!
But I see your point about how they calculate / what the 90% means, and Bermuda has eloquently summarised.
If anyone can remember the stats / normal distribution equations, we could probably calculate the percent chance of 85% or higher response rates in the whole population.
RR, it's probability theory - the greater the sample size (number of patients with results) the higher the chance it is representative of the population as a whole.
Ruck,
Have just read the rest of your last post. This is what Scancell actually said about the 90% figure:-
'The aim is to achieve at least 18 further responses (i.e., 27 responses in total) which would statistically demonstrate that SCIB1, in combination with doublet therapy, exceeds currently achievable ORRs. Recruitment is on track with data available in H1 2024. Based upon the first 13 patients there is a greater than 90% probability that the second phase will also be successful.'
So they're not saying that there's a 90% probability that the 85% will be repeated in the full 43, they're saying that there's a 90% probability that the trial will be successful ie. reach the 70% response rate threshold.
I think 3 patients were recruited to the SCIB1/Keytruda (pembrolizumab) arm of the trial. So could be 24 patients recruitied to the SCIB1/doublet CPI arm and 3 to the SCIB1/Keytruda which gives the total of 27. Or Violindog may well be right and it could be 27 recruited but only 24 dosed.
Looks like recruitment has slipped again. Originally it was meant to be complete by end of 2023 then it slipped to Q1 and now it's Q2.