You wrote (in bold):
Firstly, the MSB and the Osiris cells are different and distinct (they had to be, otherwise MSB would never have got their patent - when you read both patents, you will see how MSB clearly differentiates their cells from the Osiris cells, two totally different cell populations).
Interesting point. Sounds reasonable. Different patents for different methods. From memory Paul Simmons used the stro-1 antibody.
So the MOA might be subtly or even quite different - and in any case would need to be separately identified and established.
But I think this is not right because MSB continued to use donor cell banks (DCBs) generated by Osiris even into their most recent 54 patient pediatric GVHD001 trial. This is all based on MSB's own ODAC Briefing Document (as follows) -
MSB outlines their rationale for their proposed (MOA) Mechanism of Action in section 2 Mechanism of Action (of the ODAC Briefing Paper) which runs from page 10 to page 19.
p22 of 127 (in 4 Development History, fourth para from bottom last sentence) says -
"DCBs" (Donor Cell Banks)"manufactured by Osiris at LWI" (Lonza Walkerville Inc) "in 2008 and 2009 are intended for continued use in the production of proposed commercial DP" (Drug Product)"(see Section 7.1)".
If you look at CMC Figure 12: Summary of Manufacturing Sites used in the Product of Donor Cell Bank and Drug Product during Development to Proposed Commercial Manufacturer (on page 23) you'll see that a DCB box appears next to Osiris and the LWI manufacturing sites but NOT next to the LBSS (Lonza Bioscience Singapore) manufacturing site, which did manufacture DP (drug product). I think that makes clear no Donor Cell Banks were generated or used in MSB's trials including GVHD001 other than from the DCB's they already had.
Finally, page 33 (under section 7.1 Donor Cell Banks) is this - "Mesoblast will therefore continue to manufacture DP using DCBs already manufactured and stored at <= 135 degrees C, but will requalify each DCB lot, through retesting of defined indentity, potency, and safety attributes as an added level of assurance to ensure ongoing suitability for use prior to further manufacturing. DCBs will be retested and must meet acceptance criteria to be qualified for up to 12 months. After this period DCB's will be requalified once again to ensure ongoing suitability".
They say "will" not have been. I don't know it that is significant but it might be because had they been doing potency testing using TNFR1 expression every 12 months - they might have discovered a difference between p2 potency release tested cells at the time of the DCBs creation as they were storaged into sets of vials comprising each DCB and at the twelve monthly quality testing times for potency.
Secondly, the manufacturing process is totally different - the Osiris process is based on adhesion to plastic, the MSB process on antibody binding. So the CMC package will be majorly different.For the reasons above, same DCB's used all the way through to GVHD001 trial so the process is not different because the Donor Cell Banks weren't different - TNFR1 expression used as potency assay (p2 release test) (along with IL2Ralpha inhibition) consistently.
In MSB's announcement of July 29 2022, Mesoblast Operational and Financial Highlights for Quarter Ended June 30, 2022, MSB said -
"In response to FDA guidance, Mesoblast has optimised a potentcy assay that was in place at the time of the 54-patient Phase 3 trial in children with SR-aGVHD".
That statement is important in my opinion because its constrains what MSB can have done in terms of altering or improving the potency assay to being something to do with TNFR1 expression. (And necessarily so - they couldn't break the link with what they hope to be the pivotal trial by using potency assays that weren't measured prior to the product being put into patients, they'd not have correlations if they did that). In theory there were a range of things MSB might have done to add to the potency assay - they might have added INF gamma as a separate and additional assay (that was one used by Klinker and Bauer in their own study and its widely used in MSC research), they might have looked at other factors downstream of NF kappaB such as NF-kB phosporylation, production of M-CSF and MCP-1/CCL2 which Silviu himself says in answering a question he was asked by Dr Robey (p129 of the transcript) - Dr Robey actually asked Silviu "In going back to Dr Bauer's slide about the no clear correlation between TNFR1 levels and proposed mechanism of action, do you have any thoughts about how you can address this issue or if there are additional factors that could be used to bolster the potential mechanism of action in vivo?" .... "In other words" (DR Robey continued) "I think this is a major question that Dr Bauer has raised about the fact that there is no correlation. It's an in vitro assay, and we're concerned about what's happening in vivo. So what are your forward-looking thoughts about how you will address this issue?
Silviu answered in some detail and ended up talking about IL2Ralpa - but the slide (MA-5) Silviu and MSB put up on IL2Ralpha (several times during the presentation) is not a particularly persuasive one. And elsewhere in the ODAC presentation Silviu acknowledges that himself.
AS things stand currently, in my opinion, MSB is planning on going back to the FDA hoping to persuade on the "totality of the data" they have. That is not a different approach to what they used last time. Last time was very much a totality of the data approach too. Though to be fair they had less data then. But I think the problems are going to be similar this time in that the reason they have to go with totality of the data rather than overwhelming persuasiveness of a trial is that they still have a lot of confounding variables in their data and they need to present a number of chained correlations because in single correlations they don't have enough data.
Only 40 lots from only 3 donors were used in the hoped to be pivotal trial and some of the patients that got treated with those lots got treated from multiple lots.
They only had 25 patients with serum data that they could match up with 27 after the GVHD001 trial to compare serum levels and MAP stuff. Some of the 25 patients probably weren't patients that had just received product from a single lot.
It could be that as few as 11 single lot patients in GVHD001 had IL2Ralpha data at all. I can't think of any reason why MSB would put up a plot of % IL2Ralpha inhibition in vitro versus % change CD3+CD++CD25+(note CD25 is ILR2alpha)HLA_DR+ containing only 11 data points if they had more. (Slide MA-5). Because the R2 value of 0.38 is low and the P value of 0.04 is just significant. That slide which has "dots represent individual lots" in the bottom left is not a strong selling slide - so I assume they presented it because it was the best they had. IE. That 11 dot points representing 11 cases of patients getting lots of some measurement of IL2Ralpha inhibition was the best that they had. One of those 11 lots shows no change, no reduction of CD3+CD4+CD25+HLA-DR+ readings in serum after 27 days of treatment at all.
My point - that data MSB had to show was not powerful persuasive data. Once the confounding stuff was taken out it was too skimpy. So they had to go with a totality of the data approach back then. And it didn't work. Klinker and Bauer understood what they were being shown too well to be persuaded by connections that didn't sufficiently make for a persuasive case.
Expand
(20min delay)