Makes sense. I'm just under the impression (asking for input in case it's false) that most clinicians are responding to phase 3 trials as those are the ones that guide clinical decision more often than phase 2 results. Phase 2 results, again maybe I have this wrong, are often used to figure out safety and how effective the drug is in various areas. That's why I think treating primary endpoints and secondary endpoints in phase 2 trials as the same as in phase 3 trials isn't necessarily fair.
If the phase 2 trial shows promise on the primary but doesn't hit "significance" but quite a few meaningful secondary endpoints show "significance" and are beneficial for the patients, that would likely lead to a phase 3 trial focusing more on those areas in which significance was shown, with a shift in how the phase 3 primary endpoint was geared.
In the phase 2 trial we have here for m/m, it was initiated with collaboration from the FDA to highlight what they thought early on would be the most significant data points, and they included a slew of meaningful secondary endpoints as well to make sure they were capturing how effective the drug was, if at all. If it turns out the primary endpoint wasn't a good choice, but a fair number of secondary endpoints that are relevant and important to clinical improvement for patients, the results won't be dismissed just because they didn't hit their primary endpoint. This trial was a scattershot of endpoints to judge efficacy of leronlimab. Its efficacy doesn't rely solely on the primary endpoint, in other words, but in the bigger picture of how the drug can help people.
If the secondary endpoints that are significant can be presented in such a way to show that clinical benefit can be offered by using the drug, it'd be foolish for anyone to dismiss the trial based on the primary endpoint missing "significance", if the secondary endpoints show statistically significant efficacy in areas that are important for patient wellbeing.
All this said, i'm not really arguing with you...just making points that are up for refutation by anyone. Looking for feedback on these views.