A Normal Cortisol Day Curve, Persistent Fatigue: The Case for Modified-Release Hydrocortisone
A single patient, six years into treated Addison's disease. Her cortisol day curve is normal, so the real question isn't whether she's under-replaced by the usual test, but whether a peak-and-trough dosing schedule is producing a gap that test doesn't catch.
Marguerite T., a 41-year-old woman, has worked as an official courtroom stenographer for fourteen years — a job that gives her, better than most, a granular record of exactly when her concentration fails. Autoimmune Addison's disease was diagnosed six years ago, and she has been maintained since on hydrocortisone 10 mg–5 mg–5 mg at 7 a.m., noon, and 5 p.m., plus fludrocortisone 0.1 mg. For the past four months she has noticed a reproducible pattern rather than generalized tiredness: on trial days, her transcription accuracy — normally exact — degrades starting around 3 p.m., the same window each time, and recovers within an hour of her evening dose. She has also gained eleven pounds since her last visit without a change in diet, and her home blood pressure log shows a systolic average climbing from the 110s into the 130s over the same four months.
A three-point cortisol day curve, drawn specifically to rule out simple under-replacement before assuming the symptom is something else, came back unremarkable at all three intervals — she is not, by the test the clinic already had, inadequately dosed. That leaves the team with a harder question than a low number would have posed: whether her thrice-daily immediate-release regimen, adequate by trough-and-peak testing, is nonetheless producing exactly the kind of late-afternoon exposure gap that a peak-and-trough dosing schedule is mechanically prone to, rather than a smooth circadian profile. Johannsson et al.'s 2012 crossover trial found that a single morning dose of once-daily dual-release hydrocortisone — Plenadren, the formulation actually on the table here — produces a lower total 24-hour cortisol exposure than the same total dose split three times daily, despite a similar peak — the two schedules are not pharmacologically equivalent just because their totals match, which is exactly the gap a same-day-total day curve cannot see. She has no family history of autoimmune disease herself, though her mother developed Hashimoto's thyroiditis in her sixties, and until this year she describes her Addison's disease as something she managed without thinking about daily.
Whether the fix is a smaller change to what she is already taking or a formulation switch to modified-release hydrocortisone is complicated by one fact neither trial addresses: Plenadren is not FDA-approved and not commercially available in the United States, reachable only through a manufacturer expanded-access or named-patient import application that can itself take months to clear, with no guarantee of approval at the end of it.
Endocrinology follow-up, reviewing a normal day curve
Her day curve clears the bar we usually stop at, but I don't think it's the right bar for what she's actually reporting. Johannsson et al.'s 2012 JCEM crossover found that once-daily modified-release hydrocortisone produced a lower 24-hour cortisol AUC than the same total dose split three times daily, with lower systolic pressure and better glucose metabolism — and the DREAM trial went further, showing it reversed a measurably pro-inflammatory monocyte and natural-killer-cell profile that patients on conventional dosing carried. Her weight and blood pressure have moved in exactly the direction those trials moved in the other direction. I want to start the import application.
You're right that the metabolic drift lines up with those trials — I'm not contesting DREAM or Johannsson. What I'd push back on is treating modified-release as a general fix for a trough-timed symptom specifically. CHAMPAIN (Prete et al., 2026) put the two modified-release formulations head-to-head, and Plenadren — the one we would actually be importing — left patients waking with essentially undetectable cortisol, against several hundred nmol/L on twice-daily Chronocort. It is Chronocort, not Plenadren, that restored a physiological waking level in that trial, and Chronocort was the arm with less fatigue and better quality of life. So “modified-release” is not one thing here, and the version within reach is not the version that trial favored.
On top of that, conversion isn't free: modified-release runs roughly 20% lower bioavailability at an equivalent labeled dose, which means real under-replacement risk during any switch, layered on an import process that can take months and isn't guaranteed to clear at all. Before committing to that, I'd rather test the cheap hypothesis directly.
I don't think this actually has to be sequential. The import application's own timeline is the longest part of either path, so starting it today costs nothing and commits her to nothing — if the simpler fix works, the application just gets withdrawn later. What I'd add tonight is a small, reversible change: a 2.5 mg hydrocortisone dose at 2:30, closing the gap right where her own log says it opens, and see whether two weeks of that alone resolves what she's describing before either of you spend more time arguing about a drug she may not end up needing.
Agreed: add the 2:30 p.m. hydrocortisone micro-dose for a two-week trial against her own transcription-accuracy log, and file the modified-release import application in parallel given its independent timeline.
Not agreed: what happens if the micro-dose works. The pharmacologist sees no reason to pursue an unapproved import for a problem already solved by a cheaper, reversible change; the endocrinologist views the weight, blood pressure, and immune-profile benefits documented in DREAM as real and separate from the trough-symptom question, and worth having on their own terms even if her afternoon complaint resolves. That decision was deliberately deferred to the two-week follow-up rather than settled now.