My take is along the same lines - NEU and Acadia are a good fit.
NEU has drug for Retts that requires a hand-crafted tailored approach for P3 success because the clinical benefit from trof is modest. And Acadia has skills in getting drugs with modest clinical benefit approved.
The simplest way to understand and put a metric to “clinical benefit” is through the Clinical Global Impression – Improvement Scale (CGI-I). This scale was used in NEU’s Retts P2b trial and is the co-primary to the Retts P3 trial.
The CGI-I scale is rated (on a 7 point scale) by clinicians to measure how much patient’s condition improves (or deteriorates) during a trial from baseline. The categories are as follows:
1 Very much improved
2 Much improved
3 Minimally improved
4 No change
5 Minimally worse
6 Much worse
7 Very much worse
A full one-point CGI-I reduction compared to placebo is traditionally taken to signify clinical significance. So for example if the placebo control group on average scored 6 (much worse) and your drug group scored 5 (minimally worse) this one-point difference could be considered a clinically significant improvement.
The CGI-I was used in Arcadia’s Pimavanserin for Parkinsons Disease Psychosis (PDP) P3 trial (N~200). The trial was powered to detect a one-point reduction (representing clinical significance) on the CGI-I.
The actual average CGI-I effect found was 0.67 (p<0.01) of one-point. Conventional thinking tells you that while the improvement was statistically significant it was not clinically significant. And this is exactly why the FDA’s Medical Officer recommended against approval. But he was over turned. There were competing arguments in favour of approval.
There were currently no approved treatments for PDP and caregivers and families find PDP very upsetting and they want everything possible done for their loved ones. There was a consistent pattern of positive (modest) results across most secondary outcomes. There seemed to be a dose effect relationship. And finally a responder analysis showed quite a large positive effect for a sub-set of trial participants.
Acadia paid $250k upfront to the Ipsen group for Pimavanserin with $8m on FDA approval. Acadia funded two P2 trials and a P3 trial (all over about 10 years). There were many twists and turns – some measures worked in one trial but then failed to replicate in the next.
Considerable creativity and ingenuity was required in crafting the pattern of results achieved in P3 – including the creation of a new measure (9 items from a longer scale) of delusions which was used as the primary outcome; with the CGI-I one of the secondaries.
Pimavanserin was approved. And then the wheels fell off. Prescriptions were a fraction of what everyone had predicted. The FDA’s Chief Medical Officer’s opinion there is a not a clinically significant benefit has clout. Doctors paid “speakers fees” to present at swanky conferences promoting Pimavanserin did however prescribe quite liberally. Then safety concerns were raised which are big problem because if there is only limited (or no) clinical benefit it isn’t worth much (or even any) risk at all.
Why was Pimavanserin with Acadia in the first place and not with big pharma?
Big pharma wasn’t interested in it. Big pharma mass produces drugs for mass consumption. It tries to cut its losses in drug development as quickly as possible. Fail more efficiently than anyone else. It only takes strong efficacy signals from P2 into P3 and hopes the effects don’t shrink too much. Acadia does the opposite – the more it failed the harder it tried. Can’t make an outcome measure move? Don’t give up - invent a new measure. Solutions are hand crafted – and these solutions cannot be produced in a big pharma factory.
Turning to NEU.
In the Retts P2a trial results announcement NEU claimed statistically significant results that had met pre-specified thresholds. But it wouldn’t tell you what those thresholds were. Only long after the horse had bolted did investors learn (from the peer review paper) that NEU had used “p<0.20” and not the usual “p<0.05.
NEU created a new scale the MBAci – a subset of items from the MBA on the basis of what it thought was the strong efficacy signal (p<0.14). It agreed this with the FDA to be primary outcomes for the P3 trial. The MBAci then promptly failed to replicate in the Retts P2b trial.
No problem the RSBQ was statistically significant. And so was the CGI-I. Except that it was only 0.5 of one-point. Lower than Pimavanserin’s 0.67 (in P3) that the FDAs Medical Officer doesn't think shows clinically significance.
So what on earth are Acadia thinking?
Well they will be very much banking on the Retts IGF-1 biomarker study researchers being able to prospectively identify that sub-set of likely responders. They were the key to Pimavanserin’s approval. And that the FDA remains a political organisation where the Chief Medical Officers voice is not the final say. Acadia have done it once … so they can do it again.
Now I suppose the reason NEU can’t argue the case for Acadia in this manner is because it told investors P3 success was as easy as falling off a log. Perfect for big pharma and a big deal. The positive expectancy they created enabled the company to raise funds during the last 12 months through Landstead funding. A Claytons capital raise. Investors were the collateral damage – except it wasn’t accidental. This is the “disingenuous” bit.
All pretty stock standard … as others said it better … in junior specs the story will always be better than the outcome.
But this one did have neat twist. “In Dr T we trust”. A heuristic to reduce investor cognitive load in response to complexity. A little bit like Tomboys famous line to some very complex AHZ pre-clinical drug vaccine results “Mice don’t lie”.
The funny bit is that the “in Dr T we trust” heuristic arises because the CGI-I, which is at the root of the problem here, is actually a heuristic itself. It reduces complex judgements around what is clinically significant benefit to a simple scale score of one point or not. So you had compounding heuristics.
(20min delay)