Phase 3 Sprinter Data3 Mar 2022 21:00
I've been thinking about what might come out of the detailed analysis of the phase 3 sprinter data and I'm sure there will be quite a lot of insights that are hidden beneath the headline (binary) assessment that the end points weren't met. What I think we can reasonably say with some certainty is that there will be cohorts within the trial that responded far more to SNG than others - the headline numbers are effectively an average across the entire trial population, so within that we should expect to see quite wide variations across a range of different variables such as sex, race, known co-morbidities etc. I suspect that most of those differences will likely be either relatively marginal across a big sub-set, or less marginal across a smaller sub-set of the trial, which may make it unlikely we'll see much that has clear statistical significance at this point. There is one variable, however, that I think we should look out for in particular, as I think it will be very much on the minds of the scientists running all of these trials and that's vaccinated versus unvaccinated patients.
My reasoning is twofold:
Firstly, it's probably one of the few subdivisions of the trial population that doesn't immediately render the resulting component parts too small to provide anything meaningful - I doubt that it would be 50/50, but we know that with Omicron, the hugely increased rate of infection meant that, even with the majority of the population vaxed, there are still significant numbers of those people going into hospital. In the UK, 90% of the population is vaxed, but if just 1% of those people end up in hospital (versus e.g. 10% of unvaxed), the absolute numbers of each are pretty much the same, so that would mean that I think it's likely that the trial population contained a good sized cohort of both, which should bolster the chances of statistical significance.
Secondly, you'd have to assume that vaccination status will have had a material impact on the patient outcomes and there could therefore very easily be a material difference between the two cohorts i.e. you might see a 60% reduction in deaths for unvaxed versus a 20% reduction in vaxed (or vice versa). Could that be enough to be statistically significant across the smaller population size? I don't know, but it would definitely lend weight to undertaking a larger platform trial.
But, I hear you say, why does that difference matter when the population is being progressively vaccinated, so the numbers of unvaxed patients is continually dropping? It matters because we are just one nasty vaccine-evading variant away from the whole population becoming effectively unvaccinated again. The science suggests that this is pretty likely at some point in the future, so in that event, having a variant-agnostic therapy that's ready to go could be the difference between the world going back to square one with this virus, or having the means to manage whatever it throws at us.