Diabetes insipidus and polyuria-polydipsia — viva

Opening (must-hit)

"This is the polyuria-polydipsia syndrome: constant polyuria with polydipsia and a normal glucose, so a water diuresis. Her high-normal sodium and osmolality with dilute urine keep true diabetes insipidus in view. I need to separate three diagnoses — central diabetes insipidus, nephrogenic diabetes insipidus, and primary polydipsia — using paired osmolalities and either a water-deprivation test with a desmopressin challenge or, preferably, a copeptin-based test. Treatment then follows the mechanism: desmopressin for central disease, a low-solute diet with thiazide, amiloride and indometacin for nephrogenic disease, and fluid restriction for primary polydipsia." [1][5]

Branch A — First principles and the first test

Examiner: What is your very first investigation and why? Candidate: A glucose, to exclude osmotic diuresis from diabetes mellitus, because diabetes insipidus and diabetes mellitus share only the polyuria and take completely different paths. Once the glucose is normal, I confirm a water diuresis with paired serum and urine osmolality and a serum sodium. A high serum osmolality with an inappropriately low urine osmolality is the biochemical signature of a concentrating defect. [5]

Branch B — The vasopressin axis

Examiner: Explain the physiology I am testing here. Candidate: Hypothalamic osmoreceptors sense a rising plasma osmolality and trigger vasopressin release from the posterior pituitary. Vasopressin binds the V2 receptor on the collecting duct, and through cyclic AMP it inserts aquaporin-2 water channels into the apical membrane, so water is reabsorbed and urine is concentrated. Central diabetes insipidus breaks this at the source, nephrogenic disease breaks it at the kidney, and primary polydipsia suppresses the whole axis by drinking too much water. [1][9]

Branch C — The copeptin approach

Examiner: How would a copeptin-based approach separate the three? Candidate: Copeptin is the stable C-terminal fragment of the vasopressin precursor, released one-for-one with vasopressin, so it is a practical surrogate. A high baseline copeptin, measured without fluid restriction, identifies nephrogenic disease because vasopressin is present but ignored. To separate central disease from primary polydipsia I stimulate copeptin — with hypertonic saline to a target sodium under close monitoring, or more simply with arginine. A stimulated copeptin that stays low is central disease; an adequate rise is primary polydipsia. [7][8]

Branch D — Why not just the water-deprivation test

Examiner: Why not simply do a water-deprivation test? Candidate: It is still valid where copeptin is unavailable, but it is laborious and unreliable in exactly the situation that matters — separating partial central disease from primary polydipsia. Years of water loading in primary polydipsia wash out the medullary gradient, so the child fails to concentrate fully and looks like partial diabetes insipidus. The Fenske New England Journal of Medicine trial showed that hypertonic saline-stimulated copeptin classifies patients far more accurately than the deprivation test. [5][7]

Branch E — Treatment by mechanism

Examiner: Say she has central disease. How do you treat it, and what is the key danger? Candidate: Desmopressin, a selective V2 agonist, titrated to control the polyuria, by the oral, intranasal, or subcutaneous route — noting the subcutaneous route is far more potent, so the routes are not interchangeable. The key danger is hyponatraemia from water retention, so I deliberately allow a daily break in antidiuresis before the next dose, letting her excrete a water load. In neonates and young infants I would be especially cautious because their high fluid intake makes hyponatraemia easy to provoke. [4]

Branch F — Nephrogenic disease

Examiner: And if the urine had not concentrated with desmopressin? Candidate: That is nephrogenic disease, and desmopressin will not help. I remove reversible causes — stop lithium, correct hypercalcaemia and hypokalaemia — and reduce the solute the kidney must excrete with a low-solute, low-sodium diet and free access to water. I add a thiazide diuretic, which paradoxically reduces urine output through mild volume contraction, with amiloride (which also blocks lithium entry and spares potassium), and indometacin to potentiate concentration. [9]

Branch G — Primary polydipsia

Examiner: Suppose the copeptin rose adequately on stimulation. What now? Candidate: That is primary polydipsia — the axis is intact and she is simply drinking too much. I treat with supervised, graded fluid restriction and address the behavioural or psychiatric driver with the relevant team. I would not give desmopressin: if she keeps drinking under antidiuresis she will retain water and can develop hyponatraemic seizures. Confirming the diagnosis before treating is a safety step, not a formality. [5][9]

Branch H — The idiopathic central trap and the triphasic response

Examiner: Say she has central disease with a normal first MRI. Are you reassured? Candidate: Not fully. Idiopathic central diabetes insipidus in a child may be an occult germinoma or Langerhans cell histiocytosis that appears only later, so I measure beta-hCG and alpha-fetoprotein and arrange surveillance MRI, watching the pituitary stalk. And if she had just had pituitary surgery I would anticipate the triphasic response — diabetes insipidus, then a SIADH phase around day 5 to 10, then permanent diabetes insipidus — checking the sodium daily so I do not run desmopressin into the SIADH phase and cause hyponatraemia. [4][11]

Examiner traps

• Giving desmopressin to primary polydipsia (hyponatraemic seizures). [9]
• Correcting hypernatraemia too fast (cerebral oedema). [9]
• Accepting an idiopathic central label without surveillance (missed germinoma). [4]
• Missing adipsic diabetes insipidus, where absent thirst removes the child's protection. [11]