His Insulin Doses Were Set for a Pituitary That No Longer Works the Same Way
A single patient, longstanding type 2 diabetes on a stable insulin regimen, now three weeks out from pituitary surgery that has left him GH- and cortisol-deficient. The disagreement is how far to cut his insulin before his numbers actually prove he needs it cut.
E.S., a 66-year-old man, drove a school bus route for twenty-two years before retiring, and still keeps a printed copy of that route's schedule taped inside his kitchen cabinet, a habit his daughter finds funny and he finds necessary — the same instinct for structure that has kept his type 2 diabetes tightly logged in a notebook for the fourteen years he's managed it, most recently on 38 units of basal insulin nightly plus a sliding-scale rapid-acting dose with meals. Three weeks ago he underwent transsphenoidal resection of a 20mm nonfunctioning macroadenoma; post-operative testing confirmed new central adrenal insufficiency (now replaced on hydrocortisone 15mg) and severe GH deficiency, not yet treated, alongside preserved thyroid and gonadal function. What brought him back early, four days before his scheduled follow-up, is two mornings this week where his glucose meter read 58 and 61 before breakfast — numbers his notebook shows he has not seen in over a decade of logging, and which happened on the exact insulin dose that controlled him reliably for years before surgery.
The mechanism is not a coincidence, and it is not primarily about his new hydrocortisone dose being poorly matched — it is about what surgery removed. Growth hormone is a genuine, if often underappreciated, counter-regulatory hormone: it antagonizes insulin's action at the liver and periphery, and its sudden absence measurably increases insulin sensitivity, sometimes dramatically, in patients whose prior insulin requirements were partly calibrated against GH's own diabetogenic pull. Cortisol plays a related but distinct role, driving hepatic gluconeogenesis and opposing insulin peripherally; even physiologic-dose replacement, well below the levels his body produced before surgery removed his corticotroph reserve, restores only a fraction of that effect. Between the two losses, his actual insulin requirement has likely dropped substantially from where it sat pre-operatively, and his current dose — unchanged since before surgery, on the reasonable but now-outdated logic of "don't fix what wasn't broken" — is very plausibly delivering a relative overdose against a body that no longer generates the same resistance.
The same insulin dose, in a body that's stopped fighting it
I'd cut his basal insulin by roughly a third today, not wait for a pattern to fully establish itself. Two fasting glucoses in the 50s on an unchanged regimen, in a patient who just lost both GH and adequate cortisol reserve — two genuine counter-regulatory hormones — isn't an ambiguous signal, it's the expected consequence of exactly what surgery did to him, and waiting risks a more severe hypoglycemic event before his next visit.
I'd cut it, but I'd want a more structured basis than two fingersticks before deciding how much — a third is a reasonable starting guess, but his actual insulin sensitivity shift could be larger or smaller than that, and both his hydrocortisone dose and his eventual GH-replacement decision are both still moving targets that will keep changing his requirement over the coming months, not just this week.
Cutting by a fixed fraction today treats this like a one-time correction; it's actually the start of an ongoing recalibration that needs its own explicit monitoring plan, or we'll be having this same urgent conversation again once hydrocortisone gets adjusted or GH replacement starts.
That's the right frame, and I'd build it in now rather than treat today's cut as the fix. Reduce basal insulin by a third today given the acute hypoglycemia risk, but set structured glucose-log review at one week rather than waiting for his next scheduled visit, with an explicit note that any future hydrocortisone adjustment or start of GH replacement should trigger its own insulin reassessment, not get made in isolation.
Agreed: reduce basal insulin by one-third today, adjust the mealtime sliding scale proportionally, and schedule a structured glucose-log review at one week rather than his originally scheduled longer-interval follow-up. Explicit note added to his chart that any future hydrocortisone dose change or the start of GH replacement should independently trigger a fresh insulin reassessment, not be assumed to leave today's dose still correct.
Both physicians agreed on the acute correction; the pharmacologist's contribution was insisting the fix be framed as the first step in an ongoing recalibration tied to his other hormone decisions, not a single adjustment that resolves the issue.