See davisite imo you are fundamentally misunderstanding what...

See davisite imo you are fundamentally misunderstanding what has gone on here. You believe that when a trial fails its primary outcome it fails full stop. This is true … but only for confirmatory trials.

What you don’t understand is that trial registration sites are not designed for exploratory trials. The ANZCTR processes force you to list your efficacy outcome as a primary and to specify your supporting outcomes as secondaries.

 

Which makes perfect sense for a confirmatory trial but not an exploratory trial where the criterion for success is if you generate an interesting / important hypothesis to test in a further trial.

 

How do we know this trial was in fact exploratory and not confirmatory? Two ways.

 

First BIT state in black and white in the ASX response that the trial was exploratory.

 

Second look at the registration. The section relating to statistical methods is blank. This is an optional field because it is recognised that exploratory trials will not be powered on testing for primary and nor will they often have a pre-specified statistical analysis plan.

 

If this trial had been confirmatory it would be different in two ways. First it would have been powered with a much larger sample size to test for a possible difference in viral load. Second it would not have used this type of design. The design would be more like the treatment interruption type study BIT have proposed next.

 

Once you understand this difference between exploratory and confirmatory all your points about multiple testing, p-values etc etc just become irrelevant. What you don’t understand is really you are talking about how people go about managing the false discovery rate from exploratory studies.

 

Really you have just fallen into the same trap as the starry-eyed posters here. They think a confirmatory trial was run with positive results. You think a confirmatory trial was run with a negative result on the primary so it failed.

 

Both are wrong – an exploratory trial was run which in the opinion of BIT produced an interesting hypothesis. They saw a reduction of CD163. The p-value is completely unimportant. The effect size is what is important. BIT have provided a reference for this:

 

Knudsen, T.B., et al., Plasma Soluble CD163 Level Independently Predicts All ‐Cause Mortality in HIV‐1‐Infected Individuals. J Infect Dis, 2016. 214(8): p. 1198‐204.

 

What this forum needs is some sort of content expert who can provide a context for that CD163 reduction. Ho-hum or really exciting?

 

The difficulty here generally for ASX investors is nearly all the trials they invest in are more exploratory than confirmatory. Judging the strength of early stage efficacy signals is very tricky. The ASX reporting is terrible and to correctly interpret possible clinical significance you need some level of content expertise.

 

Some examples.

 

NEU ran a P2a Retts trial. NEU reported that the trial had met pre-specified criteria on an efficacy measure (MBA). They failed to report at the time that the pre-specified criteria was actually p<0.20. They agreed the MBA as the primary outcome with the FDA. NEU went on to run a P2b study and the MBA failed to replicate – it was ns. But another efficacy measure (RSBQ) was positive and NEU have partnered with Acardia for a P3 trial using this. The effect (if it even exists) is very modest; and the upfront from Acadia is commensurate with this. And the share price halved when the market could see exactly what its monetary value was.

 

RAP ran a US confirmatory trial – does it’s app accurately predict pneumonia. The bar was set at better than 75% sensitivity and specificity. “Better than” means here the lower bound of the 95% CI above 75%. The first US trial was a flop. The share price dropped to a third. The second US trial produced estimates all under the 75% criteria. RAP claimed it was a success for a number of other respiratory illnesses. The share price halved. RAP said investors were uneducated. RAP posters on HC are clueless the trial failed its pre-specified criteria. RAP hope to sneak a couple of the “better” indications through the FDA on the basis that a lower bar can be argued.

 

MSB … well you could write a book. After all these years, hundreds of millions of dollars, more than a dozen trials it is yet to produce a convincing confirmatory efficacy signal for anything. MSB use every trick in the book to spin their results. There is a good reason the share price sits at $1.50 miles from the $11 in its heyday.

 

The funny thing is you would cut MSB some slack for an outcome “switchoroo” but I wouldn’t because of their history. But not BIT … whereas I would … because I can I see exactly how they got in this muddle with their primary outcome.  And unlike MSB and RAP who claim investors are dumb … BIT has just got on with the show.