A term (38 weeks 2 days) female weighing 2060 g, born via spontaneous normal vaginal delivery, to a primigravida mother with gestational diabetes (on dietary modification), out of a non-consanguineous marriage, was admitted to our neonatal intensive care unit (NICU). At birth, the baby was non-vigorous (APGAR score 1 min-4/10; 5 min-7/10) and had aspirated meconium. Immediate bag and mask ventilation was started, following which the baby cried; however, she was in visible respiratory distress (Downe’s score was 5/10). She was put on bubble CPAP with appropriate settings. In view of the positive sepsis screen (Table 1) and shock (CRT > 3 s; MAP < 5th centile), empirical intravenous (IV) antibiotics (ampicillin, amikacin) and inotropes were started, respectively, along with other supportive treatments. Within 24 h, the clinical condition started improving, and there was no need for further escalation of therapy. By day 4, the baby was weaned off CPAP and was shifted on oxygen by nasal prongs; inotropes were tapered and stopped on the following day. The baby was gradually started on feeds and was tolerating well.

On the 8th post-natal day, the baby developed generalized tonic seizures followed by a dip in the alertness level. On evaluation, blood sugar (35 mg/dL) was low, whereas, ionized calcium (1.24 mmol/L), serum sodium (132 mmol/L), and potassium (4.2 mmol/L) were within the normal range. Immediate correction of hypoglycemia was done by giving a bolus of dextrose 10% along with an increase in the glucose infusion rate of the maintenance fluid. Despite these corrective measures, the baby had repeated episodes of seizures for which antiepileptics (phenobarbitone, followed by phenytoin) were started. On clinical examination, the baby was dull; anterior fontanelle was tense and bulging, with increased tone and repeated posturing episodes. Repeat blood sugar values were normal; however, seizures continued, requiring levetiracetam and midazolam infusions. Considering the possibility of meningitis, antibiotics were upgraded (meropenem) and a blood culture was sent. Baby again developed shock requiring inotropes (dopamine, adrenaline). On day 9 of life, a USG skull was planned however, it was inconclusive. Seizures were controlled on midazolam infusion (3 μg/kg/min). The baby required mannitol, acetazolamide, and 3% saline infusion for features of raised intracranial tension. In the following days, the child remained seizure-free with increased tone in all limbs, midazolam infusion was stopped on day 11. Additionally, there was a disproportionate increase in head circumference on serial monitoring (1 cm/day) with a tense anterior fontanelle and no improvement in sensorium. On day 12, a contrast-enhanced MRI brain was planned to look for the cause of seizures and altered mental status, which showed left basal ganglia bleed with intraventricular extension and watershed infarcts. MR venography revealed cerebral venous thrombosis (CSVT) involving the superior sagittal sinus, straight sinus, and transverse sinus (Fig. 1). Thrombotic workup revealed normal prothrombin and activated partial thrombin time, thrombocytopenia (107,000/mm3) and anemia (Hb 10 g/dL). Workup for inherited causes of hypercoagulable state was not possible due to financial constraints. Metabolic workup, including urine GCMS and blood TMS, was negative. Thyroid profile was normal (TSH 4.51, total T3 1.18 ng/mL, total T4 13.96 mcg/dL). Echocardiography was normal. Blood culture had grown Candida spp. for which fluconazole was added. Gradually, the child started showing some improvement in mental status and was started on orogastric feeds, followed by paladai feeds.

A T1W imaging showing left basal ganglia bleed. B MRI venography showing cerebral venous thrombosis involving superior sagittal, straight, and transverse sinus

On day 19, routine lab evaluation revealed serum sodium 129 mmol/L, which further dropped to 123 mmol/L the next day. Blood pH (7.4), serum potassium (4.3 mmol/L), and other electrolytes were within normal limits. The child also became hypotensive again, requiring fluid boluses and inotropes. Renal functions were also normal (BUN 7 mg/dL, serum creatinine 0.2 mg/dL). Considering the possibility of SIADH, a weight chart was reviewed, which, however, showed progressive weight loss over the last few days. Urine output monitoring revealed polyuria (urine output 6.5 ml/kg/h). The child also started vomiting; considering it symptomatic hyponatremia, a 3% saline bolus was given, followed by infusion at the rate of 0.1 ml/kg/h. Repeat serum sodium values after 24 h showed improvement (S Na 133 mmol/L). However, the next day serum sodium again dropped to 118 mmol/L (Fig. 2A), requiring an increase in the maintenance of 3% saline to 0.5 ml/kg/h. Over the next few days, there were repeated episodes of hyponatremia (Fig. 2A), persistent polyuria (Fig. 2B), and progressive loss of weight (Fig. 2C). In view of these findings, a possibility of renal salt wasting was considered and a urine spot sodium was sent which was 110 mmol/L (high). Serum and urine osmolality were 255 mOsm/kg (low) and 430 mOsm/kg (inappropriately high), respectively, consistent with RSW. Hypouricemia (S. urate 1.2 mg/dL) was also noted. For polyuria and weight loss, initially, extra fluids were given, followed by increased feed volume. For persistent hyponatremia, the baby was started on oral table salt 1 gm per feed three times a day (sodium 23 mmol/kg/day) on day 33 (via paladai). Thereafter, the sodium values remained within the normal range, weight gain started and the baby was shifted out of NICU. In the following days, the baby developed proper rooting, sucking, and swallowing coordination, and hence, direct breastfeeding was started. Subsequently, the baby was discharged in satisfactory clinical condition on day 38 (serum sodium 137 mmol/L, weight 2340 g) with advice to give a pinch (1 gm approx) of table salt 3 times a day mixed with expressed breast milk and review weekly.

A Serum sodium. B Urine output. C Weight in relation to day of life

On follow-up visits, the serum sodium remained within the normal range and the child was gaining weight. In view of this, sodium supplementation was gradually tapered and stopped after one and a half months. Subsequent serum sodium values remained normal.