A pregnant woman with de novo polyuria-polydipsia and elevated liver enzymes

A 32-year-old black woman, gravida 1, para 0, was admitted in our hospital in her 29th week of a gemellar pregnancy because of the risk of preterm delivery and suspicion of pre-eclampsia. The patient’s pregnancy had been uneventful until her hospitalization and no previous history of gestational diabetes or hypertension was elicited. The patient’s past medical history was unremarkable. Weight gain during pregnancy had been appropriate and serial ultrasounds had demonstrated adequate growth of both fetuses. On admission, she complained of right abdominal pain but reported no fever, vaginal bleeding or discharge, headache, blurry vision or oedema. Physical examination revealed an afebrile patient with normal skin turgor, and a tender but not enlarged liver. Her blood pressure was 100/70mmHg, and pulse 80 beats per minute. An abdominal ultrasound showed a gemellar pregnancy with normal amniotic fluid, a normal hepatic size and appearance, and a slight distension of the urinary tract. The patient was kept under observation. Laboratory values at entry showed an elevation in liver-function tests, uric acid and creatinine levels (Table 1). Haemoglobin, platelet count, coagulation and thyroid function studies, as well as potassium, calcium and glycaemia levels were within normal range. HIV as well as hepatitis serologies were negative. A urine culture remained sterile. A presumptive diagnosis of acute fatty liver of pregnancy was made. From days 2 to 9 of hospitalization, the elevation of the liver enzymes, and uric acid level persisted, but without significant exacerbation (Table 1). Creatinine clearance was 99.6ml/min. Maternal blood pressure and fetal evaluation remained normal. On day 9, the patient developed a syndrome of polydipsia and polyuria. These symptoms had been appearing the day before admission but the patient didn’t mention them at entry. A first recorded urine output was 5.0 l/24 h with a specific gravity of 1002. On day 11, urine output was 8 l, plasma osmolality 290mOsml/kg and urine osmolality 60. Since glucose, potassium and calcium levels were normal, a diagnosis of diabetes insipidus (DI) was retained. 8-D-arginin vasopressin (dDAVP) was not introduced since the course of pregnancy was marked, some hours later, by a premature rupture of the membranes with rapid labour initiation. Two healthy female babies weighing 1630 and 1650 g were delivered vaginally without complication, at 31 2/7 weeks. The patient delivered a placenta 1.5 times (640 g) the expected weight for the gestational age. Post-partum, the patient nursed her babies normally. On day 13 (1 day after delivery), the plasma level of arginin vasopressin (AVP) was 0.6 pg/ml (normal values 0.2–2.5 pg/ml). Symptoms of polyuria-polydipsia and laboratory abnormalities regressed rapidly after delivery (Table 1). Plasmatic and urinary osmolality measured 3 months after delivery, with the patient being asymptomatic, were within normal range (Table 1). The final diagnosis was transient DI of the pregnancy, probably mediated by an excessive activity of placental vasopressinase, and associated with acute fatty liver. Correspondence and offprint requests to: Frédéric Barbey, Department of Nephrology, University Hospital, Bugnon 17 avenue, 1011 Lausanne, Switzerland. Email: frederic.barbey@ chuv.hospvd.ch