Three Phases in Two Weeks: Timing Desmopressin Against a Response That Keeps Changing
A single patient, two days out from pituitary stalk manipulation during surgery, now producing four liters of dilute urine a day. The disagreement is whether to start scheduled desmopressin now or dose only as symptoms demand, given what the same injury reliably does one week from now.
K.R., a 41-year-old man, coaches a youth swim team on weekend mornings, and the detail that finally got him to mention his thirst to the covering resident, two days after his transsphenoidal surgery for a 26mm macroadenoma with stalk involvement, was that he'd gone through the entire pitcher of ice water on his tray twice before lunch and was already eyeing a third. His nursing flowsheet confirms what his own account suggested: urine output of 4.2 liters over the past 24 hours, urine specific gravity 1.002 and inappropriately dilute against a serum sodium that has climbed to 148 mEq/L — the classic biochemical signature of acute central diabetes insipidus, caused by surgical disruption of vasopressin-producing neurons in the posterior pituitary or their axonal transport down the stalk, a real and expected risk whenever stalk manipulation is required during resection.
What makes his next week genuinely harder to manage than the current picture alone suggests is the well-described triphasic pattern this specific injury produces, distinct from either simple transient or simple permanent DI. This initial phase, lasting roughly five to seven days, reflects functional shutdown of vasopressin release from stunned but not necessarily destroyed neurons. It is very often followed by a second phase, lasting several days to two weeks, in which those same neurons, now dying rather than merely stunned, release their stored vasopressin uncontrolled — producing a transient SIADH-like state with inappropriate antidiuresis and a real risk of dangerous hyponatremia, particularly in a patient still receiving desmopressin dosed for the first phase's opposite problem. Only after that unpredictable middle phase resolves does a patient's eventual, permanent DI status become clear, and roughly a third of patients who reach that point have no lasting deficit at all. Treating phase one aggressively on a fixed schedule, without a plan for phase two, is the single most common way this specific injury turns into a preventable hyponatremic emergency rather than a self-limited nuisance.
Treating the first phase without setting up a crash in the second
I'd start scheduled desmopressin now rather than waiting for symptoms to worsen further. He's already at 4.2 liters a day with a rising sodium — that's a real, current physiologic burden, and letting him keep drinking to compensate isn't sustainable indefinitely, especially overnight when he can't self-correct by drinking on his own schedule.
I'd treat this specific phase, but deliberately avoid scheduled dosing — I'd use desmopressin PRN, triggered by urine output and specific gravity thresholds rather than a fixed clock, precisely because of what comes next. This injury pattern, stalk involvement during resection, is exactly the profile most likely to produce the second, antidiuretic phase within the coming week, and a patient already on scheduled desmopressin when that phase begins is the specific setup for dangerous, hard-to-reverse hyponatremia.
Scheduled dosing solves today's problem cleanly but removes the only warning sign — breakthrough polyuria between doses — that would otherwise tell us the moment his own vasopressin release resumes and starts working against an already-dosed exogenous supply.
That's a fair point about losing the warning sign, and I'd rather have it than smoother numbers today. PRN dosing with explicit thresholds — treat when urine output exceeds 300mL/hour for two consecutive hours or specific gravity stays below 1.005 — daily sodium checks, and a standing order to hold any dose if he develops unexpected thirst suppression or falling output, which would be the earliest sign phase two has started.
Agreed: PRN desmopressin dosed against explicit urine-output and specific-gravity thresholds rather than a fixed schedule, with daily serum sodium checks through the acute window and a standing order to hold any dose if thirst or output unexpectedly drops — the earliest signal the second, antidiuretic phase has begun.
Both physicians converged on this plan once the triphasic pattern's specific risk to scheduled dosing was made explicit; the neurosurgeon's initial preference for scheduled dosing was revised once the tradeoff between smoother short-term control and losing an early warning sign was named directly.