For anyone who held MSB for throughout the years, they would have seen the progress from unsuccessful adult trials, to posthoc analysis leading to discovery of opportunity in the paediatric cohort, to most recently successful pivotal trial in the paediatric cohort which is the indication they are filing against.
Combined with some knowledge of FDA and their approach with biologics (complicated class of therapeutics, multimodal MoA, one of the toughest to approve), the questions raised by the FDA should not spook you. It is their duty to discuss all points of concern. Trials for rare diseases with poor prognosis are rarely perfect 'double blinded randomised placebo controlled trials' with perfect safety and efficacy profiles. Biologics and drugs on market without a clear understanding of the mechanism of action can be found. Infact if you read the FDA's briefing, if they really have already made a stance to deny approval based on the questions, then the historical facts and MSB's trial outlined in the very same document would have contradicted their decision. So... i don't believe a CRL is coming but further trial, hopefully one that's a post market clinical follow up i.e. after approval, is possible.
I'm not denying that there are challenges with the biologics modality in general and for MSB, the FDA's questions on critical quality attributes and clinical efficacy are fair questions. But...if you read the AM and PM brief document instead of just the questions and have done some reasonable research into MSB, you should have some of the answers already.
1. Clinical efficacy concerns
a. Concerns about the single arm design, with regards to confounders etc well look at the table below straight from the FDA briefing document and sensitivity analysis the FDA has done. Sensitivity set 1 is taking out the confounders completely from the N. gives you a better ORR! Set 2 is assuming the confounders were failed subjects - very conservative and tough assumption - ORR is still 61%. That is not quite 20 % above the 45% null hypothesis so some risks here but remember that all this is considered under the context of disease prognosis and safety as well. b. duration of response definition There are differences between FDA and MSB's definition for DOR which can be understood as how long the treatment is actively causing effect in the body. From memory i think the difference is about 1 week. The DOR is 71 days or 54 days as per MSB or FDA's definition. Slight difference, but remember that the previously approved drug "ruxolitinib" for treatment of SRaGVHD has a DOR of 0.5 month or 15 days. This is straight from the FDA briefing document.
c.They are also concerned as to whether "one statistically-positive single-arm trial, in a landscape of the multiple negative clinical trials for the treatment of aGVHD, including randomized controlled trials, is adequate to allow one to conclude that remestemcel-L is effective in the treatment of SR-aGVHD in pediatric patients." I think it is obvious that MSB will argue any trials with adult subjects are irrelevant as while it is true in trials trials/EAP 265, 270, 276 the outcome was unsuccessful, they found evidence of treatment effect in paediatric cohort in a couple of trials in table below (again, straight from the FDA briefing document.. which you all should read if you are serious about your $$$)
Also directly from the briefing "Keeping in mind the potential pitfalls of subgroups analyses and of comparing results across independent protocols, it is of interest that the Day-28 ORR is consistently 64% - 69% in remestemcel-L-treated pediatric patients with or without additional standard care salvage therapy (Table 7). "
What this means is that despite the imperfections of comparison between trials that are designed differently, remestemcel seems to consistently achieve a ORR that is ~20% around the null hypothesis used, with or without SOC. A lot of caveats but I would this that this is a positive.
d. they are also concerned about the null hypothesis and how that is set at 45% I have not looked into how the MAGIC derived null hypothesis was formulated but believe MSB has the foresight and had enough time to justify this...basics..
2. Critical Quality Attribute concerns
This is actually challenging and i think no one is 100% sure about the MoA and relationship to treatment effect, even the investigators themselves. Refer to this video from the FDA: https://www.youtube.com/watch?v=MpZSIBpJoi0
But also refer to research notes from BP and others: it is believed that there are enough evidence to demonstrate relationship between the CQAs and treatment effect. Also i made a previous post on how in the FDA's brief document the MOA does not have to be demonstrated.
Finally, remember that approval is a benefit-risk analysis: we have: - safety: no concerns - efficacy: up for discussion, pivotal trial positive, previous trial's paediatric cohort positive (but not cannot be used as evidence) - disease with huge unmet need
With a bit of luck, we should get through. i just picked up a small holding yesterday, good luck to everyone tonight
(20min delay)