That's not what phase 1 is for. Phase 1 is essentially to test "will this drug cause terrible harm to people who receive it in a reasonably short time frame?"

That's not what phase 1 is for. Phase 1 is essentially to test "will this drug cause terrible harm to people who receive it in a reasonably short time frame?" In terms of dosage guidelines, phase 1 studies often don't have variable dosing, or if they do, not an awful lot - because the question around optimal dosing is the target of phase 2. (Sometimes you can roll 2 and 3 up together, for instance, if the difference in dosing is in duration of therapy rather than an actual difference in drug administered per unit time.) Phase 1 usually has terribly small samples, because it's only looking for terribly large effects. Seeing homogeneity there, of any stripe, is neither unexpected nor problematic. (A lot of the vaccine studies rolled up phase 1/2 together because we already have a pretty good understanding of what these delivery mechanisms do to people, and so while you'd be cautious in administering to the initial study participants, it really didn't merit a "test it on five people and hold your breath" approach.)