A 48-year-old man with bipolar I disorder has been stable on lithium 750 mg daily (serum level 0.8 mEq/L) for 2 years. He presents with a 3-week history of polyuria (4–5 L/day), polydipsia, and mild cognitive slowing. His serum sodium is 148 mEq/L (normal 135–145), serum creatinine is 1.4 mg/dL (baseline 0.9), and urine osmolality is 180 mOsm/kg (normal >600 after fluid restriction). Serum lithium level is 0.9 mEq/L. What is the most appropriate next step in management?

Explanation

## Lithium-Induced Nephrogenic Diabetes Insipidus (NDI) **Key Point:** Chronic lithium use causes nephrogenic DI in 20–40% of patients. The defect is lithium-induced downregulation of aquaporin-2 (AQP2) water channels in the collecting duct, making the kidney unresponsive to ADH. First-line treatment is amiloride, a potassium-sparing diuretic that blocks lithium entry into collecting duct cells. ### Pathophysiology of Lithium-Induced NDI 1. **Lithium enters collecting duct cells** via Na^+^/K^+^-ATPase and ENaC (epithelial sodium channel) 2. **Inhibits phosphatidylinositol signalling** → reduced AQP2 expression and trafficking 3. **Result:** Collecting duct becomes unresponsive to ADH → polyuria and polydipsia 4. **Chronic effect:** Interstitial fibrosis and reduced GFR (as seen here: creatinine 1.4, up from 0.9) ### Diagnostic Criteria Met | Finding | Interpretation | |---|---| | **Polyuria (4–5 L/day)** | Exceeds normal urine output; suggests DI | | **Polydipsia** | Compensatory response to polyuria | | **Urine osmolality 180 mOsm/kg after fluid restriction** | Inappropriately low; normal is >600; confirms NDI | | **Serum sodium 148 (hypernatraemia)** | Reflects chronic free water loss | | **Serum lithium 0.9 mEq/L** | Therapeutic level; toxicity is NOT the problem | | **Elevated creatinine (1.4)** | Chronic lithium nephrotoxicity | ### Why Amiloride? ```mermaid flowchart TD A[Lithium-Induced NDI Confirmed]:::outcome --> B{Lithium Level Therapeutic?}:::decision B -->|Yes| C[Continue lithium, add amiloride]:::action B -->|No| D[Adjust lithium dose]:::action C --> E[Amiloride blocks ENaC]:::action E --> F[Reduces lithium entry into collecting duct]:::action F --> G[AQP2 expression recovers]:::action G --> H[Urine osmolality increases, polyuria resolves]:::outcome D --> I[Recheck level, then add amiloride if needed]:::action ``` **Mechanism of Amiloride:** - Blocks epithelial sodium channel (ENaC) in the collecting duct - Lithium enters cells via ENaC; blocking it reduces intracellular Li^+^ accumulation - Allows AQP2 expression to normalize - Does NOT affect lithium serum levels (amiloride is K^+^-sparing, not a diuretic that increases lithium reabsorption) **High-Yield:** Amiloride is the ONLY diuretic safe in lithium-induced NDI. Thiazides and loop diuretics WORSEN lithium toxicity by increasing renal reabsorption. **Clinical Pearl:** Polyuria from lithium-induced NDI often persists even after discontinuation; amiloride can reverse it in some patients if caught early. ### Why NOT the Other Options? | Option | Why Wrong | |---|---| | **Discontinue lithium + switch to valproate** | Lithium level is therapeutic (0.9 mEq/L) and mood is stable. NDI is a manageable side effect, not an indication to stop. Valproate has its own toxicities (hepatotoxicity, pancreatitis). | | **Reduce lithium + add thiazide** | Thiazides increase proximal tubule reabsorption of lithium → risk of toxicity. Dose reduction is unnecessary (level is therapeutic). | | **Continue lithium + prescribe DDAVP** | DDAVP is ineffective in nephrogenic DI because the collecting duct is unresponsive to ADH by definition. Amiloride addresses the underlying pathophysiology. | ### Monitoring After Starting Amiloride - Recheck serum sodium and urine osmolality in 2–4 weeks - Monitor potassium (amiloride is K^+^-sparing; risk of hyperkalaemia, especially with renal impairment) - Recheck lithium level 1 week after starting amiloride (should remain stable) - Monitor renal function (creatinine) every 6 months [cite:Kaplan & Sadock 11e Ch 31; Harrison 21e Ch 397]