Nephrogenic diabetes insipidus: the key element of paradoxical hyponatremia

Sirs, We have read with great interest the brief report of Boussemart et al. [1] published online first and strongly agree with their recommendation of strict vigilance of the fluid status in young patients with nephrogenic diabetes insipidus (NDI) who begin treatment. In fact, we would also recommend that not only are water balance and frequent clinical assessment, including hemodynamical status when necessary, crucial, but blood and urine sodium determination as well. The authors attribute their patient’s water intoxication to a strong enhanced sodium reabsorption due to the additive effect of hydrochlorothiazide and indomethacin combination therapy together with the lack of any physiological control of water intake secondary to gastric feeding. However, we experienced a water intoxication case in a 2-month-old male infant who only received initial hydrochlorothiazide treatment and was only bottle fed. Our patient was admitted to the hospital for evaluation of prolonged fever. Hypernatremia (Na 158 mEq/l) was revealed on admission, and a repeated medical history from their parents revealed that the infant had experienced polyuria–polydipsia since the first days of life, with a daily oral water intake of 500 ml added to his formula bottles, which they had considered normal. A dehydration test with subsequent 1-desamino-8-D-arginine vasopressin administration confirmed the diagnosis of NDI, and sequence analyses revealed mutation p.Trp200GlyfsX12 in exon 4 of the AVPR2 gene. Hydrochlorothiazide treatment was initiated at 2 mg/kg per day orally twice a day. Six hours after the first dose had been administered, the infant developed transient profuse sweating, paleness and hypothermia. The serum sodium concentration was 141 mEq/l at that time (pretreatment control 153 mEq/l), indicating water intoxication. Consequently, further hydrochlorothiazide treatment was stopped. The infant clinically recovered in the following hours. The treatment was reinitiated the following day at a lower dose, 1 mg/kg per day orally twice a day, which was well tolerated; no serum sodium control was obtained. The third day of treatment, 5 h after receiving the morning dose of hydrochlorothiazide, the patient's clinical condition worsened once again, with sweating, paleness, hypothermia, weakness, lethargy and gasping breathing. The serum sodium level was checked and the patient admitted to the pediatric intensive care unit (PICU). As the serum sodium concentration was 137 mEq/l, we were unable to demonstrate hyponatremia, but we hypothesized that there had probably been a significant serum sodium and subsequent osmolar shift secondary to water intoxication that had led to cerebral swelling. A cranial computerized scan showed no findings at that moment, and the patient recovered spontaneously during the following hours. Polyuria–polydipsia and hypernatremia continued the following days after the withdrawal of treatment. Therefore, 2 days later treatment was once again reinitiated at a lower dose (0.5 mg/kg per day twice a day). This treatment was well tolerated and was progressively increased during the following days (maximum 4 mg/kg per day) until neutral water balances were be achieved. Central venous catheter Pediatr Nephrol (2009) 24:2277–2278 DOI 10.1007/s00467-009-1236-4