Blastomycosis With a New Headache — Deciding Whether the Brain Is Already Involved
A single patient, whose new headache arrived before any test could confirm what it might mean. The disagreement is whether to treat the exam finding now or wait for imaging that won't change the induction regimen either way.
G.W., a 39-year-old man, has spent the past several weeks on an excavation crew cutting a new foundation trench along a riverbank — soil disruption in exactly the kind of moist, organic-rich environment blastomycosis exposure is classically tied to, though nobody made that connection until well after his cough started. He is previously healthy. Four weeks of productive cough eventually brought verrucous skin lesions on his forearm and pulmonary nodules on chest CT, both confirmed as blastomycosis on skin biopsy. What changed the conversation from routine disseminated disease to something more urgent was new this week: a severe headache unlike anything he'd had before, and one witnessed episode his wife described as him "not making sense" for close to a minute before it passed.
The guideline-driven starting point for his treatment was never really in question — under Chapman's IDSA blastomycosis guideline, disseminated disease active enough to have already produced cutaneous and pulmonary lesions meets criteria for lipid-formulation amphotericin B induction regardless of CNS status. What the new neurologic symptoms actually change is the consolidation choice waiting on the other side of induction. Itraconazole is the standard, best-studied agent for non-CNS blastomycosis, but it penetrates cerebrospinal fluid poorly — a specific, described pharmacokinetic gap, not a general caution — which is why Chapman's guideline carves out CNS disease as the one situation where a different, better-penetrating triazole — high-dose fluconazole, or voriconazole — is named for the year of follow-on therapy instead. Confirming whether his brain is actually involved, through the MRI and lumbar puncture now pending, will decide which of those two genuinely different paths his consolidation therapy takes for the next year. What no test pending today will resolve is the episode itself. A minute of incoherence witnessed by a spouse is not a finding the MRI can confirm or refute, and a clean scan would leave it exactly where it started — described, unexplained, and impossible to unsee once the year-long consolidation choice depends on it.
In the ED, before the MRI comes back
A witnessed confusional episode plus a new severe headache, in a patient with confirmed active disseminated fungal disease, is enough pretest probability to start CNS-penetrating therapy now. I don't want to wait for the MRI or LP to confirm what the exam is already suggesting — the cost of being wrong in the other direction is too high.
You're right that we shouldn't wait, but not for the reason you're giving — his disseminated disease is already active enough on its own to meet Chapman's criteria for lipid amphotericin B induction, independent of the neurologic symptoms. So the starting point converges either way. Where the CNS question actually matters is downstream, at consolidation, once induction is already underway.
That downstream choice is a real tradeoff, worth naming plainly rather than assuming one option is simply better. Itraconazole is the best-studied agent for non-CNS blastomycosis but penetrates CSF poorly — a specific pharmacokinetic gap, not a general caution — which is why Chapman's guideline routes CNS disease to high-dose fluconazole or voriconazole instead, despite fluconazole's overall efficacy against blastomycosis running somewhat behind itraconazole's for disease that doesn't involve the brain. If the MRI and LP confirm CNS involvement, I'd choose fluconazole for consolidation specifically for that penetration; if they come back clean, itraconazole remains the better-evidenced choice and I'd switch to it instead.
This isn't itraconazole being simply outperformed — it's a real efficacy cost being accepted deliberately, only if CNS involvement is actually confirmed.
Agreed: liposomal amphotericin B induction started immediately, MRI and LP proceeding as planned rather than delayed, and consolidation agent to be decided by those results — fluconazole if CNS involvement is confirmed, itraconazole if it is not.
Not agreed: whether voriconazole might have been the better CNS-directed consolidation choice over fluconazole. The pharmacologist raised it as a real open question rather than a settled alternative, and the group agreed to revisit it specifically if MRI confirms CNS disease, rather than deciding it today ahead of that result.