An Oral Anemia Drug for a Patient Ninety Minutes From the Nearest Infusion Chair
A non-dialysis CKD patient wants an oral alternative to regular injections he has real trouble reaching. Two drugs in the same new class answered the cardiovascular-safety question that matters most here — and answered it differently.
Ninety minutes separate Walter H.'s ranch from the nearest infusion chair, a fact that shaped this conversation before any drug name entered it. He is a 68-year-old man who has run cattle on the same land his father worked before him. His CKD is stage 4, eGFR 22, and his hemoglobin of 9.4 is low enough that his nephrologist raised starting an erythropoiesis-stimulating agent. Two things about him narrow the conversation before it starts. His iron stores are already adequate — ferritin 210, saturation 28% — so the easy first move, correcting iron and seeing whether the anemia follows, has nothing left to correct. And he has no known coronary disease, no prior infarct or stroke, which matters because the entire question about this drug class is cardiovascular. Walter's own first question wasn't about mechanism — it was whether he'd have to keep making that drive. He mentioned an oral pill he'd read about online, and asked directly whether that was something he could actually take instead.
That pill is a genuinely new drug class, and the two most-studied agents in it don't tell the same safety story. Daprodustat's ASCEND-ND trial tested the drug against darbepoetin specifically in non-dialysis CKD anemia — Walter's exact population — and found it noninferior for the composite cardiovascular safety endpoint, a real, direct answer to the question that matters most before handing an older man with chronic kidney disease a new mechanism of action. Vadadustat, a different molecule working through the same hypoxia-inducible-factor pathway, was tested in the same non-dialysis population by the PRO2TECT trial — and failed to meet that same noninferiority threshold for cardiovascular safety. Two drugs, one mechanism class, one population, two different answers to the one question that would decide whether either belongs in front of Walter at all.
Two drugs in one new class, one trial result apiece
Daprodustat is a real option here, not just a convenience. ASCEND-ND tested it directly against darbepoetin in exactly Walter's population — non-dialysis CKD anemia — and found it noninferior for the composite cardiovascular safety endpoint. It also happens to solve the access problem he raised first, but I wouldn't be proposing it if the cardiovascular data weren't real on its own terms.
I'd slow down before treating "HIF-PHI" as a settled, reassuring class. Vadadustat's PRO2TECT trial studied the same mechanism, the same non-dialysis population, and it failed to meet noninferiority for cardiovascular safety — a real result, not a footnote. That's a genuine divergence inside one drug class, and this is still a newer mechanism with less real-world follow-up than conventional ESAs, with open theoretical questions about off-target hypoxia-inducible-factor effects on angiogenesis and retinal vasculature that time hasn't fully answered.
I'm not saying daprodustat's result doesn't count — I'm saying one agent's trial success next to a related agent's trial failure, in the same population, is exactly the situation that calls for weighing the specific drug's own data rather than assuming the class label settles anything either direction.
That's actually the strongest argument for daprodustat specifically, read the other way. Same population, same study design, two different molecules, two different outcomes — that pattern says the safety difference tracks the specific agent, not the mechanism class as a whole. If it were a class-wide effect, we'd expect both trials to have failed the same way. They didn't. Daprodustat's own trial result is what should carry the decision for Walter, and it's a reassuring one, in his exact population.
Agreed: daprodustat started, hemoglobin and basic metabolic panel rechecked at four weeks, with Walter's own access barrier documented as a genuine clinical factor in the choice, not just a convenience.
Not agreed: the clinical pharmacologist's broader caution about the HIF-PHI mechanism's longer-term unknowns stands unresolved by today's decision — flagged for ongoing surveillance rather than treated as settled by one favorable trial result.