Invasive Aspergillosis: When the Voriconazole Level Won't Cooperate
A single patient, forty-five days post-transplant. The disagreement isn't about whether the drug is right for the disease — it's about whether his own metabolism has already made the standard dose meaningless.
T.M., a 52-year-old man, spent three decades driving long-haul freight before his AML diagnosis put him in a transplant unit bed instead of a truck cab, and forty-five days after his allogeneic stem cell transplant he is restless in a way his team has learned to read as boredom rather than distress. He was previously healthy apart from the leukemia itself. He has been on posaconazole prophylaxis since engraftment, alongside tacrolimus and mycophenolate for graft-versus-host prophylaxis, and today's CT shows a new 2.3cm right-upper-lobe nodule with a halo sign — a finding specific enough, in a patient this immunosuppressed, that the team started voriconazole before the serum galactomannan came back at 2.1, more than four times the positive threshold.
The problem surfaced two days later: his voriconazole trough came back at 0.4mcg/mL, well under the 1.0-2.0 therapeutic window, despite a dose already above standard weight-based maintenance. Voriconazole is metabolized predominantly through CYP2C19, an enzyme with well-documented functional polymorphisms, and his genotype — sent reflexively given the mismatch between dose and level — returned *17/*17, an ultra-rapid metabolizer configuration that clears the drug faster than even aggressive empiric dosing anticipated. That single result reframes the whole decision: it isn't that voriconazole failed against his aspergillosis, it's that he was never actually exposed to a treatment dose of it. Isavuconazole's clearance runs through a broader, less CYP2C19-dependent pathway, which is part of why the SECURE trial found it non-inferior to voriconazole for invasive mold disease with a materially cleaner hepatobiliary and visual side-effect profile — a finding that matters here specifically because it describes a drug whose exposure wouldn't keep moving out from under a fixed dose the way his has already, in a patient whose neutrophil count is only just beginning to recover and has little margin left to spend on a second miscalculated week.
On the transplant unit, after the genotype comes back
This is a dosing problem with a known cause, not a reason to abandon the drug with the deepest evidence base for invasive mold disease. Herbrecht's trial is still the reason voriconazole displaced amphotericin B deoxycholate as first-line therapy here, and an ultra-rapid CYP2C19 genotype is a quantifiable, correctable variable — escalate the dose under real-time TDM and confirm we've actually reached the target window before concluding the drug itself is the wrong choice.
You're right that the genotype is correctable in principle — but "real-time TDM" isn't what we actually have. Our lab turnaround on a repeat trough is 48 hours, and he's neutropenic with a visible halo sign today, not in two days. Isavuconazole's clearance doesn't route through CYP2C19 the same way; SECURE found it non-inferior to voriconazole for invasive mold disease with meaningfully less hepatobiliary and visual toxicity, and it gives us a drug whose exposure isn't still drifting out from under whatever dose we pick next.
Escalating voriconazole again means dosing blind for two more days in a patient whose last dose already missed target by more than half — that's not a monitored correction, it's a second guess with the same lag built in.
Whichever triazole we land on, his tacrolimus needs to move today, not after a level comes back high. Isavuconazole is a weaker CYP3A4 inhibitor than voriconazole but it isn't a non-interacting drug — I'd cut his tacrolimus dose by roughly half at the switch and recheck in 72 hours rather than waiting for a rejection-range trough to tell us we were slow. Worth naming plainly: posaconazole already failed him as prophylaxis, so whatever we choose now is a second real decision, not a default extension of the first one.
Agreed within the hour: isavuconazole started at the standard loading regimen, tacrolimus dose cut by half with a recheck at 72 hours, and a repeat CT/galactomannan planned at day 7 to confirm the switch is working rather than assuming it.
Not agreed: whether isavuconazole itself needs routine TDM going forward. It isn't standard practice the way voriconazole monitoring is, and the pharmacologist would still want a level given how badly the last drug's assumed exposure diverged from its actual one; the ID physician sees the whole point of the switch as not needing that same surveillance burden. No level was ordered today, but the pharmacologist's objection is documented, not overridden.