Idiopathic Anaphylaxis: Empiric Prophylaxis or Continued Trigger Search
A high school biology teacher has had four anaphylactic episodes in fourteen months with no identified trigger. The disagreement isn't whether she needs help now — it's whether starting standing medication should wait on a workup that hasn't yet captured the one test result that could still change everything.
J.M., a 39-year-old woman who has taught high school biology for eleven years, drew her own epinephrine for the fourth time in fourteen months two weeks ago — this time at her daughter's soccer game, with no meal, medication, or exertion she can point to as the trigger. The first episode followed a restaurant dinner and looked food-related until skin testing to the meal's actual ingredients came back negative; the second happened at her desk between classes; the third while she was folding laundry at home, alone until her husband found her already reaching for the injector. Between episodes she is entirely well — an avid runner who has kept up her three-mile mornings without incident, no urticaria pigmentosa on exam, nothing suggesting a skin-level mast cell disease. A baseline tryptase drawn at a follow-up visit, weeks after her most recent episode, came back at 4.2 ng/mL, unremarkable. What has changed is not her health but her confidence: she no longer drives her children anywhere without her husband in the car, and last month she asked the school nurse to keep a spare injector in her office, a small accommodation that says more about the cost of unpredictability than any single lab value does.
Idiopathic anaphylaxis is a diagnosis reached by exclusion, and the exclusion so far has been genuinely thorough — extended food and venom skin testing, a careful medication and supplement review, no exercise or temperature pattern across four separate episodes. Greenberger's frequency-based classification of idiopathic anaphylaxis, developed out of Patterson's Northwestern series and the framework most allergists use to decide how aggressively to treat while a trigger search continues, reserves its standing recommendation for a prophylactic glucocorticoid taper for patients with six or more episodes a year or two within two months — a threshold her four episodes across fourteen months do not actually clear, even though four episodes in that span is by any reasonable measure a lot to live through. The tryptase drawn on her chart is also not quite the number the question needs: an acute-phase tryptase drawn within the first hour or two of a reaction, and paired against a true symptom-free baseline taken at least a day afterward, can reveal a transient mediator surge that neither sample shows alone; hers has never been captured that way, because none of her four episodes happened near a lab.
At the bedside
Four visits to my department in fourteen months, and the median time from her first symptom to me watching her push the injector herself is measured in minutes, not enough time to ask what she ate for lunch, let alone wait on a genetic test result. I want her on standing daily therapy today — a high-dose non-sedating H1 blocker plus an H2 blocker, at minimum — because the number that worries me isn't her tryptase, it's four.
I recognize four in fourteen months is below whatever frequency threshold makes this an easy call on paper. It doesn't feel that far below it from where I sit.
I don't dispute the frequency, I dispute treating it as settled. Greenberger's own frequency criteria for a standing prophylactic taper are written for six or more episodes a year, or two within two months — her pattern, real and frightening as it is, sits under that bar, and treating her as if it's already been cleared skips a workup step that's still genuinely open. None of her four tryptase measurements were drawn acutely; every one was a baseline value weeks removed from an episode, which is exactly the sample least likely to catch a transient mediator surge. I want a plan in her hand for her next episode — draw tryptase within the first hour or two of symptom onset, then a true baseline at least twenty-four hours after everything has settled, since the consensus comparison is an acute value exceeding 1.2 times her own baseline plus 2 — plus hereditary alpha-tryptasemia genetic testing and alpha-gal specific IgE now, because a positive result on either would change this conversation entirely, from lifelong empiric drugs to a nameable, and in HaT's case actionable, diagnosis.
I'm not asking her to wait through another unmedicated episode to get that data — I'm asking that the testing not get quietly dropped once medication starts, which is what tends to happen once a patient feels better.
I don't think this is actually a sequencing problem. A second-generation H1 antihistamine and an H2 blocker together carry close to no signal for harm at standard doses, and starting them today doesn't touch a single one of the tests just described — not the acute tryptase, not the HaT panel, not the alpha-gal IgE. The instinct to finish the workup before treating makes sense when the treatment carries real risk or the testing is slow enough that untreated time is the greater harm; neither applies to a first-line antihistamine bridge ordered alongside labs drawn the same visit. Reserve the harder call — daily glucocorticoids, omalizumab — for if she has a breakthrough episode on the antihistamine alone. That's the only decision actually worth deferring.
Agreed within the visit: start cetirizine 20mg and famotidine 20mg twice daily today; hereditary alpha-tryptasemia genetic testing and alpha-gal specific IgE ordered the same visit; she leaves with a laminated card instructing her to have tryptase drawn within the first hour or two of any future episode, with the paired baseline drawn at a follow-up visit at least twenty-four hours after symptoms have fully resolved.
Not agreed, and left open rather than smoothed over:
The allergist wants glucocorticoids reserved for a truly refractory pattern given long-term steroid toxicity.
The emergency medicine physician leans toward omalizumab sooner given her occupation and inability to predict episodes at all.