Okay, I can’t prove it. What I can say is that the #30 slide on Dr. B’s presentation at ASCO stated that the work on the SAP began back 5 months ago, which from end of May (as June would not count because the month had just started), dates back to January 2019.
And NWBO did not receive their Sawston money until late December. So it makes sense somewhat that the SAP work didn’t begin in earnest until January 2019 because the slide states that.
But that’s when likely the SAP began, and not necessarily the data collection, which one could assume has been ongoing.
The data collation and scrub is on a different track, although the CRO must be paid for their analysis and NWBO was pretty cash poor around November/December 2018 until the Sawston sale money came through. So I think I combined the two tracks in my mind, incorrectly.
Let’s look back at the 2018 data presented at SNO on November 16, 2016. One of LL’s slides indicated that the data presented dated back to October 31, 2018 or thereabouts.
Now that could be accurate, I’d like to think it’s accurate, but I don’t know if it’s accurate. If it was, that means that the blinded data they used for that SNO presentation was gleaned from data obtained 15 days previously. So October 31, 2018 would have been the soft data lock for the data presented about two weeks later.
Backing up further, DI spoke to hopeforthefuture2 and Hope walks away thinking that the soft data lock might have occurred in January 2019 and that the data gathered previously to that will count as the primary analysis data. Additionally, he thinks then that perhaps all the data that follows will not count in the primary analysis. This is because DI is describing a “lag” between the soft data lock and the hard data lock (that will date back to the soft data lock date) and he uses January 2019 as an example. So I further link the 5 months Dr. B had on his slide, convinced that there was a soft data lock that had been January 2019. And you know what, maybe that was, at one time, a time the company had thought might be their soft data lock. Or maybe it wasn’t.
Regardless of that, I think when the data is ready to be locked as the primary analysis data it is dependent upon the point in the time that a CRO is able to determine the data they have on hand thus far is fully accurate to. In other words, the CRO is asked at what point in time is the current collation of data existing reliable as a set of data that can be hard locked. For example, if one were to ask the CRO if the data were reliable enough to hard lock it dated today, it is doubtful that the CRO would respond with a yes to that question. Instead, the CRO would have to look back at the data accumulated thus far and determine at what point in time would the data be accurate and complete to for a data lock to be set to. Would it be two weeks such as might be the case if the slide on LL’s presentation of 10/31/18 were accurate, and then they analyzed the blinded data and she presented it two weeks later? Or might it be a month or two or more that they could effectively go back to and be certain the data is accurate?
So there is a gap or a lag between when the data is hard-locked (and this date cannot be changed and after that takes place, it is very, very difficult to make changes to that data) and when the soft-lock occurs. This is likely the point DI was talking to and it’s a bit difficult, I think, to fully communicate.
I think then, that the soft-lock date could conceivably change and progress forward, if it were to be of benefit to the primary analysis data. And it’s certainly conceivable it would be, given that the long tail in this trial is key; and so if the primary analysis data could to be extended to include data past October 31, 2018 or January 2019, and the CRO confirmed that it could be brought forward to a later date, then perhaps the data is somewhat of a “moveable feast” as the soft data lock moves forward.
So… I was likely incorrect in firmly thinking that January 2019 was a hard date for the soft data lock. Maybe that’s why it’s called a soft data lock. I guess then that I am moving towards your mindset, and perhaps the soft data lock may not have even been determined yet, although if not, then when it is, it would date back to where the CRO thinks it is safe to establish that date.
Anyhow, that’s my best understanding of the process after looking at it more closely. Perhaps I should have done that before stepping in on your discussion, but in doing so, I now think I understand the process a bit more clearly. There will be a lag between the two data points (which is what hope was also stating), and once the soft lock is firmly established (although it seems it is adjustable - which is good because that favors the long tail data), it will lock in what will comprise the primary analysis (which is I believe what you were stating). When the hard lock takes place, it will date back to the official date of the soft lock - whatever that is, be it two weeks, two months, or six months. And that is the date the CRO can comfortable state all the data is clean of all blemishes. Of course, they continue collecting data, but all data after the hard lock is officially in place will be “further analysis” data.
And the more money a company has, the larger staff a company can throw at such a project, the smaller the gap might be between the two - the hard data lock and the soft data lock - because the company with more of both can speed up the process because a larger staff can accumulate more data more quickly than a smaller one. Then again, with all of the blinded interim looks over the multiple years, it could be that the tortoise will have caught up to the hare because they’ve been accumulating and scrubbing this data for soooo long.
In summary then, perhaps I was dating the soft lock back to January 2019 based on what Dr. Bosch indicated on his slide pertaining to the SAP. And instead, that soft lock may be softly moving forward as you instead thought.
News 
Market Data 
Discover 
